October 2015 - lvm2-commits - Fedora mailing-lists

master - post-release
by Alasdair Kergon 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=7831a650914f35c9… Commit: 7831a650914f35c97f9deab6b871970b303a8948 Parent: 15a97cc610712efe818f96b9630b72de89a0cf57 Author: Alasdair G Kergon <agk(a)redhat.com> AuthorDate: Sat Oct 31 01:35:47 2015 +0000 Committer: Alasdair G Kergon <agk(a)redhat.com> CommitterDate: Sat Oct 31 01:35:47 2015 +0000 post-release --- VERSION | 2 +- VERSION_DM | 2 +- WHATS_NEW | 3 +++ WHATS_NEW_DM | 3 +++ 4 files changed, 8 insertions(+), 2 deletions(-) diff --git a/VERSION b/VERSION index 20d1cc4..f36c01d 100644 --- a/VERSION +++ b/VERSION @@ -1 +1 @@ -2.02.133(2)-git (2015-10-30) +2.02.134(2)-git (2015-10-30) diff --git a/VERSION_DM b/VERSION_DM index ef5a8f8..ebca82f 100644 --- a/VERSION_DM +++ b/VERSION_DM @@ -1 +1 @@ -1.02.110-git (2015-10-30) +1.02.111-git (2015-10-30) diff --git a/WHATS_NEW b/WHATS_NEW index cfd0a1a..7431483 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,3 +1,6 @@ +Version 2.02.134 - +==================================== + Version 2.02.133 - 30th October 2015 ==================================== Support repeated -o|--options for reporting commands. diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 5004f84..113a4bb 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,3 +1,6 @@ +Version 1.02.111 - +==================================== + Version 1.02.110 - 30th October 2015 ==================================== Disable thin monitoring plugin when it fails too often (>10 times).

1 0

v2_02_133 annotated tag has been created
by Alasdair Kergon 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=af330ff842aef4d4… Commit: af330ff842aef4d473f7dd9e17eb7a066f373528 Parent: 0000000000000000000000000000000000000000 Author: Alasdair G Kergon <agk(a)redhat.com> AuthorDate: 2015-10-30 15:35 +0000 Committer: Alasdair G Kergon <agk(a)redhat.com> CommitterDate: 2015-10-30 15:35 +0000 annotated tag: v2_02_133 has been created at af330ff842aef4d473f7dd9e17eb7a066f373528 (tag) tagging 15a97cc610712efe818f96b9630b72de89a0cf57 (commit) replaces v2_02_132 Release 2.02.133. This is a development release that provides us with a checkpoint. (Any interface changes in this release might be modified further.) 329 files changed, 6524 insertions(+), 4152 deletions(-) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.14 (GNU/Linux) iEYEABECAAYFAlYzj1YACgkQIoGRwVZ+LBc0rwCeI1s7/WqxlO4EDpeVuAfZvk4x LyEAniwQZ5fUTo+VBx48GfmqvJHUukOb =M010 -----END PGP SIGNATURE----- Alasdair G Kergon (15): post-release segtypes: Use SEG_TYPE_NAME_ string constants. segtypes: Make constants ULL. style: lv_manip.c changes segtypes: Add and use new segtype macros. stripes: Introduce _round_to_stripe_boundary. alloc: Move _calc_area_multiple. test: Cope with stripe rounding message change. libdm: Move status fns from deptree to targets. raid: Fail if kernel status denominator is zero. raid: Add metadata dev information to reports. Revert "log: no file for external logging" thin: Fix typo in policy threshold message. device: Tidy DASD CDL format detection code. pre-release David Teigland (26): man lvmthin: use --poolmetadatasize option lockd: fix changing lock type on VG with pool lockd: fix rescanning VG metadata lockd: add debug logging for metadata error lockd: add VG name to debug logging config: improve cache_settings description man lvmcache: updates for mode and policy lockd: add error message for EEXIST vgcreate: allow both --shared and --lock-type options vgcreate: improve error message for multiple lock managers man lvmsystemd: fix typo man lvmsystemid: fix typos Fix segfault when lvmlockd is running but not lvmetad lvmlockd: use flag to avoid blocking in sanlock_acquire pvscan: use process_each_pv libdaemon: allow all debug levels for stderr libdaemon: fix typo in last log enable commit man: document possible lvmetad -l values vg_read: sometimes ignore read errors vgextend: fix use of the wrong flag vg_read: remove unused inconsistent check pvs: fix missing PVs when VG is removed toollib: fix wrong paren placement vgs, lvs: ignore error if VG is removed pvs: don't treat duplicate PVs as missed New entries for pvs related fixes. Heinz Mauelshagen (11): segtypes: Extend flags to 64 bits. style: Miscellaneous tidying up of metadata/lv* libdm: Improve raid segment parameter handling. raid: Fix raid target write_behind parameter. libdm: fix bogus macro causing false parameter count raid: Detect whether or not kernel supports raid0. raid: Add a segtype flag for each raid type. segtypes: Use flags in raid segtype macros. raid: Use loop for text uint32_t parameter import. raid: Introduce DEFAULT_RAID_MAX_IMAGES raid_manip: fix wrong image size allocation on raid10 "lvconvert --replace ..." Lidong Zhong (1): pvcreate: don't support unpartitioned DASD devices with CDL formatted Marian Csontos (2): test: Fix timeout handling cleanup: snapshots of snapshots message Ondrej Kozina (6): polldaemon: fix missing mirror in-sync info with lvmpolld tests: add test for missing mirror in-sync info lvmetad: add optional timeout option lvmetad.8.in: update man page with timeout info tests: add lvmetad shutdown on idle test tests: replace invalid use of 'fail' with 'die' Peter Rajnoha (37): systemd: use {local,remote}-fs-pre.target instead of {local,remote}-fs.target blkdeactivate: recognize and deactivate MD devices too report: identify LV hodling sanlock locks as 'private,lockd,sanlock' within lv_role report field metadata: format_text: provide more detailed error message when metadata too large for PV mda report: add vg_missing_pv_count field vgimportclone: cleanup the script and remove dependency on awk, grep, cut and tr vgimportclone: fix vgimportclone with -n to not add number unnecessarily to base VG name vgimportclone: make sure there's no prefix in VG name list tests: test vgimportclone and -n option wiping: make libblkid detect all copies of the same signature if use_blkid_wiping=1 libdm: add dm_report_compact_given_fields config: add report/compact_output_cols to control which columns to compact in report output cleanup: typo in comment: compact_output_fields should be compact-output_cols conf: regenerate conf: clarify compact output settings pvremove: make sure even invalid info is removed from lvmcache on pvremove lvmcache: update cached info properly when moving from VG to orphan while lvmetad is used revert: 3d03e504cda6307f692c5879f6850b2e27af7d71: message about VG metadata size vs. PV mda size metadata: format_text: check metadata area size is at least MDA_SIZE_MIN metadata: format_text: check VG metadata do not overlap themselves WHATS_NEW: recent commits metadata: format_text: better check for metadata overlap tests: add test for minimum mda size metadata: format_text: also count with calculated mda size of 0 str_list: also allow memory allocation without memory pool str_list: add str_list_to_str and str_to_str_list functions str_list: add str_list_destroy function refactor: move code detecting report options to a separate function report: make report options defined by "-o" groupable report: support "-o -field_name1,field_name2,...." report: add report_get_field_prefix function report: recognize known prefix when processing "-o -field_name1,field_name2,..." report: support "-o #field_name1,field_name2,..." tests: add report-fields.sh test man: pvs/vgs/lvs and -o+, -o-, -o# commands: update command help for -o|--options for reporting commands WHATS_NEW: reporting commands and -o-/-o+ extension Zdenek Kabelac (118): alloc: fix update or area_len man: update dmsetup and dmstats pages man: update lvm pages tests: use proper target name for check tests: ensure test will not block cleanup: fix compiler warning man: fsadm man: lvcreate continue tests: check devices with prefix aren't in use libdm: enforce writethrough mode for cleaner thin: check for overflown pool earlier dmeventd: correct ret code for fail thin: enhance logging toolcontext: do not change already set dm_log log: no file for external logging dmeventd: fix loging report dmeventd: log all commands with CMD: prefix libdmevent: add dm_event_log dmeventd: syslog -> log_xxxxx dmeventd: use log_ macros in plugins dmeventd: enable new logging dmeventd: enable new logging dmeventd: drop changing logging dmeventd: add more debug messages man: dmeventd update dmeventd: handle snapshot overflow cleanup: static with _ dmeventd: support logging on stdout cleanup: more readable code cleanup: always set nsec cleanup: remove multilog cleanup: avoid using private cleanup: move system defines before structs cleanup: typo in comment dmeventd: new initialization of plugin threads dmeventd: introduce macro for init dmeventd: use new macros to init mempool dmeventd: add local mempool for raid and mirror dmeventd: less locking for snapshot and thin dmeventd: less locking mirror dmeventd: rework raid plugin makefiles: generate Local makefiles: drop explicit linking makefiles: src dirs have depth 5 tests: show status in in_sync tests: better regex check tests: start logging dmeventd output tests: wait for initial sync tests: more extend testing dmeventd: compilable without DEBUG CFLAG makefiles: older gcc needs hint with rpath dmeventd: fix missing '!' makefiles: switch to rpath-link thin: enforce local activation when creation new thin dmeventd: use matching function lib: better reporting of threshold memlock: report memlock daemon counter tools: preselect some setting for memlocked daemons tools: do not change signals for memlocked daemons libdm: add dm_hold_control_dev dmeventd: check for malloc return dmeventd: lvm2 plugin correctly debug dmeventd: raid plugin reporting dmeventd: mirror plugin update dmeventd: snapshot plugin updates dmeventd: snapshot plugin device removal dmeventd: thin plugin update dmeventd: no registering of 0 event mask dmeventd: using warning level dmeventd: move dso handling into single code section dmeventd: use dm_hold_control_dev dmeventd: wake up timer when setting new timeout dmeventd: drop unneded test dmeventd: minimize locking time for get_imeout dmeventd: code mode _get_device_status cleanup: use enums dmeventd: rework locking code dmeventd: handle signal from plugin dmeventd: snapshot plugin unmonitor dmeventd: debug signals dmeventd: exit on idle cleanup: declare vars before code cleanup: easier to read code cleanup: drop uneeded header file cleanup: indents tests: enable raid test on 4.3 dmeventd: debug error paths dmeventd: ensure filter is empty tools: update for lvm2_disable_dmeventd_monitoring cleanup: gcc warning for old-style cleanup: fix gcc compile with older pthread cleanup: drop debug thin: fix percentage compare lv_manip: ensure it will fit bellow threshold libdm: dm_tree_node_size_changed recognizes reduction dev_manager: enabled no_flush for suspend dev_manager: no flush for extension libdm: enable no_flush for driver version > 11 tests: snapshot now respects threshold dmevent: fix referencing thin: limit no-flush using only for thin-pool tests: extend timer for 4 hours tests: support skip flags tests: use skip flags dmeventd: event string parser handles empty field tests: avoid reading utils when skipping tests: use more SKIP tests: using matching type thin: dmeventd plugin check number of failures thin: enable usage of kernel low_water_mark cleanup: easier error messages cleanup: use same type cleanup: remove thin low_water_mark from metadata cleanup: error is not a WARNING tests: no point in using should tests: fix wrong line has been commented tests: let pass bigger readahead tests: update test for resize

1 0

master - pre-release
by Alasdair Kergon 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=15a97cc610712efe… Commit: 15a97cc610712efe818f96b9630b72de89a0cf57 Parent: 3f1c63c812ba5f79768f23b8d6d2e2965a458e0e Author: Alasdair G Kergon <agk(a)redhat.com> AuthorDate: Fri Oct 30 15:34:00 2015 +0000 Committer: Alasdair G Kergon <agk(a)redhat.com> CommitterDate: Fri Oct 30 15:34:00 2015 +0000 pre-release --- VERSION | 2 +- VERSION_DM | 2 +- WHATS_NEW | 4 ++-- WHATS_NEW_DM | 4 ++-- test/shell/snapshot-remove-dmsetup.sh | 2 +- 5 files changed, 7 insertions(+), 7 deletions(-) diff --git a/VERSION b/VERSION index 3d59279..20d1cc4 100644 --- a/VERSION +++ b/VERSION @@ -1 +1 @@ -2.02.133(2)-git (2015-09-22) +2.02.133(2)-git (2015-10-30) diff --git a/VERSION_DM b/VERSION_DM index 0cc8062..ef5a8f8 100644 --- a/VERSION_DM +++ b/VERSION_DM @@ -1 +1 @@ -1.02.110-git (2015-09-22) +1.02.110-git (2015-10-30) diff --git a/WHATS_NEW b/WHATS_NEW index 3808cb6..cfd0a1a 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,5 @@ -Version 2.02.133 - -====================================== +Version 2.02.133 - 30th October 2015 +==================================== Support repeated -o|--options for reporting commands. Support -o- and -o# for reporting commands to remove and compact fields. Fix missing PVs from pvs output if vgremove is run concurrently. diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 63cb2cc..5004f84 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,5 @@ -Version 1.02.110 - -====================================== +Version 1.02.110 - 30th October 2015 +==================================== Disable thin monitoring plugin when it fails too often (>10 times). Fix/restore parsing of empty field '-' when processing dmeventd event. Enhance dm_tree_node_size_changed() to recognize size reduction. diff --git a/test/shell/snapshot-remove-dmsetup.sh b/test/shell/snapshot-remove-dmsetup.sh index 6a4f9d2..14023a0 100644 --- a/test/shell/snapshot-remove-dmsetup.sh +++ b/test/shell/snapshot-remove-dmsetup.sh @@ -36,7 +36,7 @@ mount -o errors=remount-ro "$DM_DEV_DIR/$vg1/snap" mnt sync -# intentionally suspend layer bellow +# intentionally suspend layer below dmsetup suspend $vg-$lv1 # now this should pass without blocking

1 0

master - WHATS_NEW: reporting commands and -o-/-o+ extension
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=3f1c63c812ba5f79… Commit: 3f1c63c812ba5f79768f23b8d6d2e2965a458e0e Parent: dd52721b68a7c393a54149649f518f4e2d9440b5 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 30 15:50:35 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:50:35 2015 +0100 WHATS_NEW: reporting commands and -o-/-o+ extension --- WHATS_NEW | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 415efc7..3808cb6 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,7 @@ Version 2.02.133 - ====================================== + Support repeated -o|--options for reporting commands. + Support -o- and -o# for reporting commands to remove and compact fields. Fix missing PVs from pvs output if vgremove is run concurrently. Remove unwanted error message when running pvs/vgs/lvs and vgremove at once. Check newly created VG's metadata do not overlap in metadata ring buffer.

1 0

master - commands: update command help for -o|--options for reporting commands
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=dd52721b68a7c393… Commit: dd52721b68a7c393a54149649f518f4e2d9440b5 Parent: 7f125c1116df74d23b442304d8109a784f677251 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 30 15:43:57 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 commands: update command help for -o|--options for reporting commands --- tools/commands.h | 14 +++++++------- 1 files changed, 7 insertions(+), 7 deletions(-) diff --git a/tools/commands.h b/tools/commands.h index ae5dff7..a141135 100644 --- a/tools/commands.h +++ b/tools/commands.h @@ -70,7 +70,7 @@ xx(devtypes, "\t[--nameprefixes]\n" "\t[--noheadings]\n" "\t[--nosuffix]\n" - "\t[-o|--options [+]Field[,Field]]\n" + "\t[-o|--options [+|-|#]Field[,Field]]\n" "\t[-O|--sort [+|-]key1[,[+|-]key2[,...]]]\n" "\t[--rows]\n" "\t[-S|--select Selection]\n" @@ -431,7 +431,7 @@ xx(lvdisplay, "\t[--ignoreskippedcluster]\n" "\t[--noheadings]\n" "\t[--nosuffix]\n" - "\t[-o|--options [+]Field[,Field]]\n" + "\t[-o|--options [+|-|#]Field[,Field]]\n" "\t[-O|--sort [+|-]key1[,[+|-]key2[,...]]]\n" "\t[-S|--select Selection]\n" "\t[-P|--partial]\n" @@ -665,7 +665,7 @@ xx(lvs, "\t[--nameprefixes]\n" "\t[--noheadings]\n" "\t[--nosuffix]\n" - "\t[-o|--options [+]Field[,Field]]\n" + "\t[-o|--options [+|-|#]Field[,Field]]\n" "\t[-O|--sort [+|-]key1[,[+|-]key2[,...]]]\n" "\t[-P|--partial]\n" "\t[--readonly]\n" @@ -846,7 +846,7 @@ xx(pvdisplay, "\t[--ignoreskippedcluster]\n" "\t[--noheadings]\n" "\t[--nosuffix]\n" - "\t[-o|--options [+]Field[,Field]]\n" + "\t[-o|--options [+|-|#]Field[,Field]]\n" "\t[-O|--sort [+|-]key1[,[+|-]key2[,...]]]\n" "\t[-S|--select Selection]\n" "\t[--readonly]\n" @@ -938,7 +938,7 @@ xx(pvs, "\t[--nameprefixes]\n" "\t[--noheadings]\n" "\t[--nosuffix]\n" - "\t[-o|--options [+]Field[,Field]]\n" + "\t[-o|--options [+|-|#]Field[,Field]]\n" "\t[-O|--sort [+|-]key1[,[+|-]key2[,...]]]\n" "\t[-P|--partial]\n" "\t[--readonly]\n" @@ -1177,7 +1177,7 @@ xx(vgdisplay, "\t[--ignoreskippedcluster]\n" "\t[--noheadings]\n" "\t[--nosuffix]\n" - "\t[-o|--options [+]Field[,Field]]\n" + "\t[-o|--options [+|-|#]Field[,Field]]\n" "\t[-O|--sort [+|-]key1[,[+|-]key2[,...]]]\n" "\t[-P|--partial]\n" "\t[-S|--select Selection]\n" @@ -1350,7 +1350,7 @@ xx(vgs, "\t[--nameprefixes]\n" "\t[--noheadings]\n" "\t[--nosuffix]\n" - "\t[-o|--options [+]Field[,Field]]\n" + "\t[-o|--options [+|-|#]Field[,Field]]\n" "\t[-O|--sort [+|-]key1[,[+|-]key2[,...]]]\n" "\t[-P|--partial]\n" "\t[--readonly]\n"

1 0

master - man: pvs/vgs/lvs and -o+, -o-, -o#
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=7f125c1116df74d2… Commit: 7f125c1116df74d23b442304d8109a784f677251 Parent: 5b04eda93f903d25e25385d148d28b8cfa1157ea Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 30 15:36:35 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 man: pvs/vgs/lvs and -o+, -o-, -o# --- man/lvdisplay.8.in | 2 +- man/lvs.8.in | 11 ++++++++--- man/pvdisplay.8.in | 2 +- man/pvs.8.in | 11 ++++++++--- man/vgdisplay.8.in | 2 +- man/vgs.8.in | 11 ++++++++--- 6 files changed, 27 insertions(+), 12 deletions(-) diff --git a/man/lvdisplay.8.in b/man/lvdisplay.8.in index ed1d3f7..f14fa03 100644 --- a/man/lvdisplay.8.in +++ b/man/lvdisplay.8.in @@ -37,7 +37,7 @@ lvdisplay \(em display attributes of a logical volume .RB [ \-\-noheadings ] .RB [ \-\-nosuffix ] .RB [ \-o | \-\-options -.RI [ + ] Field [ ,Field ...]] +.RI [ + | \- | # ] Field [ ,Field ...]] .RB [ \-O | \-\-sort .RI [ + | \- ] Key1 [ , [ + | \- ] Key2 ...]] .RB [ \-P | \-\-partial ] diff --git a/man/lvs.8.in b/man/lvs.8.in index 3cb2ec0..952b6ec 100644 --- a/man/lvs.8.in +++ b/man/lvs.8.in @@ -16,7 +16,7 @@ lvs \(em report information about logical volumes .RB [ \-\-noheadings ] .RB [ \-\-nosuffix ] .RB [ \-o | \-\-options -.RI [ + ] Field [, Field ]] +.RI [ + | \- | # ] Field [, Field ]] .RB [ \-O | \-\-sort .RI [ + | \- ] Key1 [,[ + | \- ] Key2 [,...]]] .RB [ \-P | \-\-partial ] @@ -76,8 +76,13 @@ Suppress the suffix on output sizes. Use with \fB\-\-units\fP (except h and H) if processing the output. .TP .BR \-o ", " \-\-options -Comma-separated ordered list of columns. Precede the list with '\fI+\fP' -to append to the default selection of columns instead of replacing it. +Comma-separated ordered list of columns. +.IP +Precede the list with '\fI+\fP' to append to the current list +of columns, '\fI-\fP' to remove from the current list of columns +or '\fI#\fP' to compact given columns. The \fI\-o\fP option can +be repeated, providing several lists. These lists are evaluated +from left to right. .IP Use \fB\-o lv_all\fP to select all logical volume columns, and \fB\-o seg_all\fP diff --git a/man/pvdisplay.8.in b/man/pvdisplay.8.in index c1232d6..5f93634 100644 --- a/man/pvdisplay.8.in +++ b/man/pvdisplay.8.in @@ -37,7 +37,7 @@ pvdisplay \- display attributes of a physical volume .RB [ \-\-noheadings ] .RB [ \-\-nosuffix ] .RB [ \-o | \-\-options -.RI [ + ] Field [ ,Field ...]] +.RI [ + | \- | # ] Field [ ,Field ...]] .RB [ \-O | \-\-sort .RI [ + | \- ] Key1 [ , [ + | \- ] Key2 ... .RI ]] diff --git a/man/pvs.8.in b/man/pvs.8.in index 605b13d..efafbde 100644 --- a/man/pvs.8.in +++ b/man/pvs.8.in @@ -16,7 +16,7 @@ pvs \(em report information about physical volumes .RB [ \-\-noheadings ] .RB [ \-\-nosuffix ] .RB [ \-o | \-\-options -.RI [ + ] Field [ ,Field ...]] +.RI [ + | \- | # ] Field [ ,Field ...]] .RB [ \-O | \-\-sort .RI [ + | \- ] Key1 [ , [ + | \- ] Key2 ...]] .RB [ \-P | \-\-partial ] @@ -65,8 +65,13 @@ Suppress the suffix on output sizes. Use with \fB\-\-units\fP (except h and H) if processing the output. .TP .BR \-o ", " \-\-options -Comma-separated ordered list of columns. Precede the list with '\fI+\fP' -to append to the default selection of columns. +Comma-separated ordered list of columns. +.IP +Precede the list with '\fI+\fP' to append to the current list +of columns, '\fI-\fP' to remove from the current list of columns +or '\fI#\fP' to compact given columns. The \fI\-o\fP option can +be repeated, providing several lists. These lists are evaluated +from left to right. .IP Use \fB-o pv_all\fP to select all physical volume columns, and \fB-o pvseg_all\fP to select all Physical Volume segment columns. diff --git a/man/vgdisplay.8.in b/man/vgdisplay.8.in index 172dba4..89a7a5a 100644 --- a/man/vgdisplay.8.in +++ b/man/vgdisplay.8.in @@ -37,7 +37,7 @@ vgdisplay \(em display attributes of volume groups .RB [ \-\-noheadings ] .RB [ \-\-nosuffix ] .RB [ \-o|\-\-options -.RI [ + ] Field1 [ ,Field2 ...]] +.RI [ + | \- | # ] Field1 [ ,Field2 ...]] .RB [ \-O | \-\-sort .RI [ + | \- ] Key1 [ , [ + | \- ] Key2 ...]] .RB [ \-P | \-\-partial ] diff --git a/man/vgs.8.in b/man/vgs.8.in index 654e8bb..f219665 100644 --- a/man/vgs.8.in +++ b/man/vgs.8.in @@ -16,7 +16,7 @@ vgs \(em report information about volume groups .RB [ \-\-noheadings ] .RB [ \-\-nosuffix ] .RB [ \-o | \-\-options -.RI [ + ] Field1 [ ,Field2 ...]] +.RI [ + | \- | # ] Field1 [ ,Field2 ...]] .RB [ \-O | \-\-sort .RI [ + | \- ] Key1 [ , [ + | \- ] Key2 ...]] .RB [ \-P | \-\-partial ] @@ -63,8 +63,13 @@ Suppress the suffix on output sizes. Use with \fB\-\-units\fP (except h and H) if processing the output. .TP .BR \-o ", " \-\-options -Comma-separated ordered list of columns. Precede the list with '+' to append -to the default selection of columns. +Comma-separated ordered list of columns. +.IP +Precede the list with '\fI+\fP' to append to the current list +of columns, '\fI-\fP' to remove from the current list of columns +or '\fI#\fP' to compact given columns. The \fI\-o\fP option can +be repeated, providing several lists. These lists are evaluated +from left to right. .IP Use \fB\-o vg_all\fP to select all volume group columns. .IP

1 0

master - tests: add report-fields.sh test
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=5b04eda93f903d25… Commit: 5b04eda93f903d25e25385d148d28b8cfa1157ea Parent: 77c31d0c39c3ba9caf7a5a70282f1c5a6b0f48e2 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 30 15:06:49 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 tests: add report-fields.sh test --- test/shell/report-fields.sh | 91 +++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 91 insertions(+), 0 deletions(-) diff --git a/test/shell/report-fields.sh b/test/shell/report-fields.sh new file mode 100644 index 0000000..e0c7308 --- /dev/null +++ b/test/shell/report-fields.sh @@ -0,0 +1,91 @@ +#!/bin/sh +# Copyright (C) 2015 Red Hat, Inc. All rights reserved. +# +# This copyrighted material is made available to anyone wishing to use, +# modify, copy, or redistribute it subject to the terms and conditions +# of the GNU General Public License v.2. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software Foundation, +# Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + +SKIP_WITH_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 +SKIP_WITH_CLVMD=1 + +. lib/inittest + +# Test only that there's correct set of fields displayed on output. + +aux prepare_pvs 1 + +OPTS="--nameprefixes --noheadings --rows" + +aux lvmconf 'report/pvs_cols="pv_name,pv_size"' +aux lvmconf 'report/compact_output=0' +aux lvmconf 'report/compact_output_cols=""' + +pvs $OPTS > out +grep LVM2_PV_NAME out +grep LVM2_PV_SIZE out + +pvs $OPTS -o pv_attr > out +grep LVM2_PV_ATTR out +not grep -v LVM2_PV_ATTR out + +pvs $OPTS -o+pv_attr > out +grep LVM2_PV_NAME out +grep LVM2_PV_SIZE out +grep LVM2_PV_ATTR out + +pvs $OPTS -o-pv_name > out +not grep LVM2_PV_NAME out +grep LVM2_PV_SIZE out + +pvs $OPTS -o+pv_attr -o-pv_attr > out +grep LVM2_PV_NAME out +grep LVM2_PV_SIZE out +not grep LVM2_PV_ATTR out + +pvs $OPTS -o-pv_attr -o+pv_attr > out +grep LVM2_PV_NAME out +grep LVM2_PV_SIZE out +grep LVM2_PV_ATTR out + +pvs $OPTS -o+pv_attr -o-pv_attr -o pv_attr > out +grep LVM2_PV_ATTR out +not grep -v LVM2_PV_ATTR out + +# -o-size is the same as -o-pv_size - the prefix is recognized +pvs $OPTS -o-size > out +not grep LVM2_PV_SIZE out + +# PV does not have tags nor is it exported if we haven't done that explicitly. +# Check compaction per field is done correctly. +pvs $OPTS -o pv_name,pv_exported,pv_tags -o#pv_tags > out +grep LVM2_PV_NAME out +grep LVM2_PV_EXPORTED out +not grep LVM2_PV_TAGS out + +aux lvmconf 'report/compact_output_cols="pv_tags"' + +pvs $OPTS -o pv_name,pv_exported,pv_tags > out +grep LVM2_PV_NAME out +grep LVM2_PV_EXPORTED out +not grep LVM2_PV_TAGS out + +pvs $OPTS -o pv_name,pv_exported,pv_tags -o#pv_exported > out +grep LVM2_PV_NAME out +not grep LVM2_PV_EXPORTED out +grep LVM2_PV_TAGS out + +aux lvmconf 'report/compact_output=1' +pvs $OPTS -o pv_name,pv_exported,pv_tags > out +grep LVM2_PV_NAME out +not grep LVM2_PV_EXPORTED out +not grep LVM2_PV_TAGS out + +pvs $OPTS -o pv_name,pv_exported,pv_tags -o#pv_exported > out +grep LVM2_PV_NAME out +not grep LVM2_PV_EXPORTED out +not grep LVM2_PV_TAGS out

1 0

master - report: support "-o #field_name1, field_name2, ..."
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=77c31d0c39c3ba9c… Commit: 77c31d0c39c3ba9caf7a5a70282f1c5a6b0f48e2 Parent: baf320455b06141eb472faa687e7fcdc838bdd8c Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Thu Oct 22 15:37:27 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 report: support "-o #field_name1,field_name2,..." --- tools/reporter.c | 34 +++++++++++++++++++++++++++++----- 1 files changed, 29 insertions(+), 5 deletions(-) diff --git a/tools/reporter.c b/tools/reporter.c index f4f9b71..89024c2 100644 --- a/tools/reporter.c +++ b/tools/reporter.c @@ -642,17 +642,21 @@ static void _del_option_from_list(struct dm_list *sll, const char *prefix, static int _get_report_options(struct cmd_context *cmd, report_type_t report_type, - const char **options) + const char **options, + const char **fields_to_compact) { const char *prefix = report_get_field_prefix(report_type); size_t prefix_len = strlen(prefix); struct arg_value_group_list *current_group; struct dm_list *final_opts_list; + struct dm_list *final_compact_list = NULL; struct dm_list *opts_list = NULL; + int opts_list_destroy = 1; struct dm_str_list *sl; const char *opts; int r = ECMD_PROCESSED; + if (!(final_opts_list = str_to_str_list(NULL, *options, ",", 1))) { r = ECMD_FAILED; goto_out; @@ -673,6 +677,8 @@ static int _get_report_options(struct cmd_context *cmd, case '+': /* fall through */ case '-': + /* fall through */ + case '#': if (!(opts_list = str_to_str_list(NULL, opts + 1, ",", 1))) { r = ECMD_FAILED; goto_out; @@ -683,8 +689,17 @@ static int _get_report_options(struct cmd_context *cmd, dm_list_iterate_items(sl, opts_list) _del_option_from_list(final_opts_list, prefix, prefix_len, sl->str); + } else if (*opts == '#') { + if (!final_compact_list) { + final_compact_list = opts_list; + opts_list_destroy = 0; + } else + dm_list_splice(final_compact_list, opts_list); } - str_list_destroy(opts_list, 1); + if (opts_list_destroy) + str_list_destroy(opts_list, 1); + else + opts_list_destroy = 1; opts_list = NULL; break; default: @@ -700,9 +715,17 @@ static int _get_report_options(struct cmd_context *cmd, r = ECMD_FAILED; goto out; } + if (final_compact_list && + !(*fields_to_compact = str_list_to_str(cmd->mem, final_compact_list, ","))) { + dm_pool_free(cmd->mem, (char *) *options); + r = ECMD_FAILED; + goto out; + } out: if (opts_list) str_list_destroy(final_opts_list, 1); + if (final_compact_list) + str_list_destroy(final_compact_list, 1); if (final_opts_list) str_list_destroy(final_opts_list, 1); return r; @@ -720,7 +743,7 @@ static int _report(struct cmd_context *cmd, int argc, char **argv, unsigned args_are_pvs; int lv_info_needed, lv_segment_status_needed; int lock_global = 0; - const char *fields_to_compact; + const char *fields_to_compact = NULL; aligned = find_config_tree_bool(cmd, report_aligned_CFG, NULL); buffered = find_config_tree_bool(cmd, report_buffered_CFG, NULL); @@ -792,7 +815,7 @@ static int _report(struct cmd_context *cmd, int argc, char **argv, /* If -o supplied use it, else use default for report_type */ if (arg_count(cmd, options_ARG) && - ((r = _get_report_options(cmd, report_type, &options) != ECMD_PROCESSED))) + ((r = _get_report_options(cmd, report_type, &options, &fields_to_compact) != ECMD_PROCESSED))) return r; /* -O overrides default sort settings */ @@ -909,7 +932,8 @@ static int _report(struct cmd_context *cmd, int argc, char **argv, if (find_config_tree_bool(cmd, report_compact_output_CFG, NULL)) { if (!dm_report_compact_fields(report_handle)) log_error("Failed to compact report output."); - } else if ((fields_to_compact = find_config_tree_str_allow_empty(cmd, report_compact_output_cols_CFG, NULL))) { + } else if (fields_to_compact || + (fields_to_compact = find_config_tree_str_allow_empty(cmd, report_compact_output_cols_CFG, NULL))) { if (!dm_report_compact_given_fields(report_handle, fields_to_compact)) log_error("Failed to compact given columns in report output."); }

1 0

master - report: recognize known prefix when processing "-o -field_name1, field_name2, ..."
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=baf320455b06141e… Commit: baf320455b06141eb472faa687e7fcdc838bdd8c Parent: bb4d3fa7a70881b72ddf61cba02760b89cf07b53 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Thu Oct 22 14:30:11 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 report: recognize known prefix when processing "-o -field_name1,field_name2,..." --- tools/reporter.c | 35 +++++++++++++++++++++++++++++++---- 1 files changed, 31 insertions(+), 4 deletions(-) diff --git a/tools/reporter.c b/tools/reporter.c index 591f4ed..f4f9b71 100644 --- a/tools/reporter.c +++ b/tools/reporter.c @@ -604,22 +604,48 @@ static void _check_pv_list(struct cmd_context *cmd, int argc, char **argv, } } -static void _del_option_from_list(struct dm_list *sll, const char *str) +static void _del_option_from_list(struct dm_list *sll, const char *prefix, + size_t prefix_len, const char *str) { struct dm_list *slh; struct dm_str_list *sl; + const char *a = str, *b; + prefix_len--; dm_list_uniterate(slh, sll, sll) { sl = dm_list_item(slh, struct dm_str_list); + + /* exact match */ if (!strcmp(str, sl->str)) { dm_list_del(slh); return; } + + /* also try to match with known prefix */ + b = sl->str; + if (!strncmp(prefix, a, prefix_len)) { + a += prefix_len; + if (*a == '_') + a++; + } + if (!strncmp(prefix, b, prefix_len)) { + b += prefix_len; + if (*b == '_') + b++; + } + if (!strcmp(a, b)) { + dm_list_del(slh); + return; + } } } -static int _get_report_options(struct cmd_context *cmd, const char **options) +static int _get_report_options(struct cmd_context *cmd, + report_type_t report_type, + const char **options) { + const char *prefix = report_get_field_prefix(report_type); + size_t prefix_len = strlen(prefix); struct arg_value_group_list *current_group; struct dm_list *final_opts_list; struct dm_list *opts_list = NULL; @@ -655,7 +681,8 @@ static int _get_report_options(struct cmd_context *cmd, const char **options) dm_list_splice(final_opts_list, opts_list); } else if (*opts == '-') { dm_list_iterate_items(sl, opts_list) - _del_option_from_list(final_opts_list, sl->str); + _del_option_from_list(final_opts_list, prefix, + prefix_len, sl->str); } str_list_destroy(opts_list, 1); opts_list = NULL; @@ -765,7 +792,7 @@ static int _report(struct cmd_context *cmd, int argc, char **argv, /* If -o supplied use it, else use default for report_type */ if (arg_count(cmd, options_ARG) && - ((r = _get_report_options(cmd, &options) != ECMD_PROCESSED))) + ((r = _get_report_options(cmd, report_type, &options) != ECMD_PROCESSED))) return r; /* -O overrides default sort settings */

1 0

master - report: add report_get_field_prefix function
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=bb4d3fa7a70881b7… Commit: bb4d3fa7a70881b72ddf61cba02760b89cf07b53 Parent: 3e18b101a0845f826de29cfd0c554c5e2be2baa5 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Thu Oct 22 13:27:59 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 report: add report_get_field_prefix function --- lib/report/report.c | 15 +++++++++++++++ lib/report/report.h | 1 + 2 files changed, 16 insertions(+), 0 deletions(-) diff --git a/lib/report/report.c b/lib/report/report.c index dea66ee..911c96e 100644 --- a/lib/report/report.c +++ b/lib/report/report.c @@ -3396,6 +3396,21 @@ void *report_init_for_selection(struct cmd_context *cmd, cmd); } +const char *report_get_field_prefix(report_type_t report_type_id) +{ + const struct dm_report_object_type *report_types, *report_type; + + report_types = report_type_id & DEVTYPES ? _devtypes_report_types + : _report_types; + + for (report_type = report_types; report_type->id; report_type++) { + if (report_type_id & report_type->id) + return report_type->prefix; + } + + return ""; +} + /* * Create a row of data for an object */ diff --git a/lib/report/report.h b/lib/report/report.h index 84c5f91..6c8b026 100644 --- a/lib/report/report.h +++ b/lib/report/report.h @@ -71,6 +71,7 @@ void *report_init(struct cmd_context *cmd, const char *format, const char *keys, int quoted, int columns_as_rows, const char *selection); void *report_init_for_selection(struct cmd_context *cmd, report_type_t *report_type, const char *selection); +const char *report_get_field_prefix(report_type_t report_type); int report_for_selection(struct cmd_context *cmd, struct selection_handle *sh, struct physical_volume *pv,

1 0

master - report: support "-o -field_name1, field_name2, ...."
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=3e18b101a0845f82… Commit: 3e18b101a0845f826de29cfd0c554c5e2be2baa5 Parent: df190dcfa58ee3c7cc9959da6af3ae22e9618992 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Wed Oct 21 15:36:26 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 report: support "-o -field_name1,field_name2,...." --- tools/reporter.c | 24 +++++++++++++++++++++++- 1 files changed, 23 insertions(+), 1 deletions(-) diff --git a/tools/reporter.c b/tools/reporter.c index f053b08..591f4ed 100644 --- a/tools/reporter.c +++ b/tools/reporter.c @@ -604,11 +604,26 @@ static void _check_pv_list(struct cmd_context *cmd, int argc, char **argv, } } +static void _del_option_from_list(struct dm_list *sll, const char *str) +{ + struct dm_list *slh; + struct dm_str_list *sl; + + dm_list_uniterate(slh, sll, sll) { + sl = dm_list_item(slh, struct dm_str_list); + if (!strcmp(str, sl->str)) { + dm_list_del(slh); + return; + } + } +} + static int _get_report_options(struct cmd_context *cmd, const char **options) { struct arg_value_group_list *current_group; struct dm_list *final_opts_list; struct dm_list *opts_list = NULL; + struct dm_str_list *sl; const char *opts; int r = ECMD_PROCESSED; @@ -630,11 +645,18 @@ static int _get_report_options(struct cmd_context *cmd, const char **options) switch (*opts) { case '+': + /* fall through */ + case '-': if (!(opts_list = str_to_str_list(NULL, opts + 1, ",", 1))) { r = ECMD_FAILED; goto_out; } - dm_list_splice(final_opts_list, opts_list); + if (*opts == '+') { + dm_list_splice(final_opts_list, opts_list); + } else if (*opts == '-') { + dm_list_iterate_items(sl, opts_list) + _del_option_from_list(final_opts_list, sl->str); + } str_list_destroy(opts_list, 1); opts_list = NULL; break;

1 0

master - report: make report options defined by "-o" groupable
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=df190dcfa58ee3c7… Commit: df190dcfa58ee3c7cc9959da6af3ae22e9618992 Parent: e149fe7fdfd22b496bbd870d19412b0e9f090149 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Wed Oct 21 14:57:49 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 report: make report options defined by "-o" groupable Also, besides making "-o" groupable, use string lists to store lists of options temporarily while processing all instances of the "-o" group. --- tools/args.h | 2 +- tools/reporter.c | 61 ++++++++++++++++++++++++++++++++++++++++------------- 2 files changed, 47 insertions(+), 16 deletions(-) diff --git a/tools/args.h b/tools/args.h index c887774..210545e 100644 --- a/tools/args.h +++ b/tools/args.h @@ -185,7 +185,7 @@ arg(name_ARG, 'n', "name", string_arg, 0) arg(nofsck_ARG, 'n', "nofsck", NULL, 0) arg(novolumegroup_ARG, 'n', "novolumegroup", NULL, 0) arg(oldpath_ARG, 'n', "oldpath", NULL, 0) -arg(options_ARG, 'o', "options", string_arg, 0) +arg(options_ARG, 'o', "options", string_arg, ARG_GROUPABLE) arg(sort_ARG, 'O', "sort", string_arg, 0) arg(maxphysicalvolumes_ARG, 'p', "maxphysicalvolumes", int_arg, 0) arg(permission_ARG, 'p', "permission", permission_arg, 0) diff --git a/tools/reporter.c b/tools/reporter.c index 4cf099b..f053b08 100644 --- a/tools/reporter.c +++ b/tools/reporter.c @@ -606,26 +606,57 @@ static void _check_pv_list(struct cmd_context *cmd, int argc, char **argv, static int _get_report_options(struct cmd_context *cmd, const char **options) { + struct arg_value_group_list *current_group; + struct dm_list *final_opts_list; + struct dm_list *opts_list = NULL; const char *opts; - char *str; + int r = ECMD_PROCESSED; - opts = arg_str_value(cmd, options_ARG, ""); - if (!opts || !*opts) { - log_error("Invalid options string: %s", opts); - return EINVALID_CMD_LINE; + if (!(final_opts_list = str_to_str_list(NULL, *options, ",", 1))) { + r = ECMD_FAILED; + goto_out; } - if (*opts == '+') { - if (!(str = dm_pool_alloc(cmd->mem, - strlen(*options) + strlen(opts) + 1))) { - log_error("options string allocation failed"); - return ECMD_FAILED; + + dm_list_iterate_items(current_group, &cmd->arg_value_groups) { + if (!grouped_arg_is_set(current_group->arg_values, options_ARG)) + continue; + + opts = grouped_arg_str_value(current_group->arg_values, options_ARG, NULL); + if (!opts || !*opts) { + log_error("Invalid options string: %s", opts); + r = EINVALID_CMD_LINE; + goto out; } - (void) sprintf(str, "%s,%s", *options, opts + 1); - *options = str; - } else - *options = opts; - return ECMD_PROCESSED; + switch (*opts) { + case '+': + if (!(opts_list = str_to_str_list(NULL, opts + 1, ",", 1))) { + r = ECMD_FAILED; + goto_out; + } + dm_list_splice(final_opts_list, opts_list); + str_list_destroy(opts_list, 1); + opts_list = NULL; + break; + default: + str_list_destroy(final_opts_list, 1); + if (!(final_opts_list = str_to_str_list(NULL, opts, ",", 1))) { + r = ECMD_FAILED; + goto out; + } + } + } + + if (!(*options = str_list_to_str(cmd->mem, final_opts_list, ","))) { + r = ECMD_FAILED; + goto out; + } +out: + if (opts_list) + str_list_destroy(final_opts_list, 1); + if (final_opts_list) + str_list_destroy(final_opts_list, 1); + return r; } static int _report(struct cmd_context *cmd, int argc, char **argv,

1 0

master - refactor: move code detecting report options to a separate function
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=e149fe7fdfd22b49… Commit: e149fe7fdfd22b496bbd870d19412b0e9f090149 Parent: 77605457e7cf550364355b03cb5544b57d3c046d Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Tue Oct 20 16:36:11 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 refactor: move code detecting report options to a separate function --- tools/reporter.c | 46 +++++++++++++++++++++++++++------------------- 1 files changed, 27 insertions(+), 19 deletions(-) diff --git a/tools/reporter.c b/tools/reporter.c index 6c93739..4cf099b 100644 --- a/tools/reporter.c +++ b/tools/reporter.c @@ -604,13 +604,35 @@ static void _check_pv_list(struct cmd_context *cmd, int argc, char **argv, } } +static int _get_report_options(struct cmd_context *cmd, const char **options) +{ + const char *opts; + char *str; + + opts = arg_str_value(cmd, options_ARG, ""); + if (!opts || !*opts) { + log_error("Invalid options string: %s", opts); + return EINVALID_CMD_LINE; + } + if (*opts == '+') { + if (!(str = dm_pool_alloc(cmd->mem, + strlen(*options) + strlen(opts) + 1))) { + log_error("options string allocation failed"); + return ECMD_FAILED; + } + (void) sprintf(str, "%s,%s", *options, opts + 1); + *options = str; + } else + *options = opts; + + return ECMD_PROCESSED; +} + static int _report(struct cmd_context *cmd, int argc, char **argv, report_type_t report_type) { void *report_handle; struct processing_handle handle = {0}; - const char *opts; - char *str; const char *keys = NULL, *options = NULL, *selection = NULL, *separator; int r = ECMD_PROCESSED; int aligned, buffered, headings, field_prefixes, quoted; @@ -689,23 +711,9 @@ static int _report(struct cmd_context *cmd, int argc, char **argv, } /* If -o supplied use it, else use default for report_type */ - if (arg_count(cmd, options_ARG)) { - opts = arg_str_value(cmd, options_ARG, ""); - if (!opts || !*opts) { - log_error("Invalid options string: %s", opts); - return EINVALID_CMD_LINE; - } - if (*opts == '+') { - if (!(str = dm_pool_alloc(cmd->mem, - strlen(options) + strlen(opts) + 1))) { - log_error("options string allocation failed"); - return ECMD_FAILED; - } - (void) sprintf(str, "%s,%s", options, opts + 1); - options = str; - } else - options = opts; - } + if (arg_count(cmd, options_ARG) && + ((r = _get_report_options(cmd, &options) != ECMD_PROCESSED))) + return r; /* -O overrides default sort settings */ keys = arg_str_value(cmd, sort_ARG, keys);

1 0

master - str_list: add str_list_destroy function
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=77605457e7cf5503… Commit: 77605457e7cf550364355b03cb5544b57d3c046d Parent: 0d5b1294f0350e10698232eb764c28b3ca54df1a Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Tue Oct 20 16:12:16 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 str_list: add str_list_destroy function The str_list_destroy function may be called to cleanup memory when the list is not used anymore and the list itself was not allocated from the memory pool. --- lib/datastruct/str_list.c | 13 +++++++++++++ lib/datastruct/str_list.h | 2 ++ 2 files changed, 15 insertions(+), 0 deletions(-) diff --git a/lib/datastruct/str_list.c b/lib/datastruct/str_list.c index a600f60..73b5534 100644 --- a/lib/datastruct/str_list.c +++ b/lib/datastruct/str_list.c @@ -247,3 +247,16 @@ bad: dm_pool_free(mem, list); return NULL; } + +void str_list_destroy(struct dm_list *list, int deallocate_strings) +{ + struct dm_str_list *sl, *tmp_sl; + + dm_list_iterate_items_safe(sl, tmp_sl, list) { + dm_list_del(&sl->list); + if (deallocate_strings) + dm_free((char *)sl->str); + dm_free(sl); + } + dm_free(list); +} diff --git a/lib/datastruct/str_list.h b/lib/datastruct/str_list.h index 268a3cf..92ff8c5 100644 --- a/lib/datastruct/str_list.h +++ b/lib/datastruct/str_list.h @@ -32,5 +32,7 @@ int str_list_dup(struct dm_pool *mem, struct dm_list *sllnew, const struct dm_list *sllold); char *str_list_to_str(struct dm_pool *mem, const struct dm_list *list, const char *delim); struct dm_list *str_to_str_list(struct dm_pool *mem, const char *str, const char *delim, int ignore_multiple_delim); +/* Only for lists which were *not* allocated from the mem pool! */ +void str_list_destroy(struct dm_list *list, int deallocate_strings); #endif

1 0

master - str_list: add str_list_to_str and str_to_str_list functions
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=0d5b1294f0350e10… Commit: 0d5b1294f0350e10698232eb764c28b3ca54df1a Parent: 097d14e64e845bdb1c0b8106e33f38f775e54fe5 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Tue Oct 20 16:01:10 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:56 2015 +0100 str_list: add str_list_to_str and str_to_str_list functions The str_list_to_str and str_to_str_list are helper functions to convert string list to a single string and vice versa. --- lib/datastruct/str_list.c | 87 +++++++++++++++++++++++++++++++++++++++++++++ lib/datastruct/str_list.h | 2 + 2 files changed, 89 insertions(+), 0 deletions(-) diff --git a/lib/datastruct/str_list.c b/lib/datastruct/str_list.c index 25ffddd..a600f60 100644 --- a/lib/datastruct/str_list.c +++ b/lib/datastruct/str_list.c @@ -16,6 +16,8 @@ #include "lib.h" #include "str_list.h" +#include <ctype.h> + struct dm_list *str_list_create(struct dm_pool *mem) { struct dm_list *sl; @@ -160,3 +162,88 @@ int str_list_lists_equal(const struct dm_list *sll, const struct dm_list *sll2) return 1; } + +char *str_list_to_str(struct dm_pool *mem, const struct dm_list *list, + const char *delim) +{ + size_t delim_len = strlen(delim); + unsigned list_size = dm_list_size(list); + struct dm_str_list *sl; + char *str, *p; + size_t len = 0; + unsigned i = 0; + + dm_list_iterate_items(sl, list) + len += strlen(sl->str); + if (list_size > 1) + len += ((list_size - 1) * delim_len); + + str = mem ? dm_pool_alloc(mem, len+1) : dm_malloc(len+1); + if (!str) { + log_error("str_list_to_str: string allocation failed."); + return NULL; + } + str[len] = '\0'; + p = str; + + dm_list_iterate_items(sl, list) { + len = strlen(sl->str); + memcpy(p, sl->str, len); + p += len; + + if (++i != list_size) { + memcpy(p, delim, delim_len); + p += delim_len; + } + } + + return str; +} + +struct dm_list *str_to_str_list(struct dm_pool *mem, const char *str, + const char *delim, int ignore_multiple_delim) +{ + size_t delim_len = strlen(delim); + struct dm_list *list; + const char *p1, *p2, *next; + char *str_item; + size_t len; + + if (!(list = str_list_create(mem))) { + log_error("str_to_str_list: string list allocation failed."); + return NULL; + } + + p1 = p2 = str; + while (*p1) { + if (!(p2 = strstr(p1, delim))) + next = p2 = str + strlen(str); + else + next = p2 + delim_len; + + len = p2 - p1; + str_item = mem ? dm_pool_alloc(mem, len+1) : dm_malloc(len+1); + if (!str_item) { + log_error("str_to_str_list: string list item allocation failed."); + goto bad; + } + memcpy(str_item, p1, len); + str_item[len] = '\0'; + + if (!str_list_add_no_dup_check(mem, list, str_item)) + goto_bad; + + if (ignore_multiple_delim) { + while (!strncmp(next, delim, delim_len)) + next += delim_len; + } + + p1 = next; + } + + return list; +bad: + if (mem) + dm_pool_free(mem, list); + return NULL; +} diff --git a/lib/datastruct/str_list.h b/lib/datastruct/str_list.h index 3121a28..268a3cf 100644 --- a/lib/datastruct/str_list.h +++ b/lib/datastruct/str_list.h @@ -30,5 +30,7 @@ int str_list_match_list(const struct dm_list *sll, const struct dm_list *sll2, c int str_list_lists_equal(const struct dm_list *sll, const struct dm_list *sll2); int str_list_dup(struct dm_pool *mem, struct dm_list *sllnew, const struct dm_list *sllold); +char *str_list_to_str(struct dm_pool *mem, const struct dm_list *list, const char *delim); +struct dm_list *str_to_str_list(struct dm_pool *mem, const char *str, const char *delim, int ignore_multiple_delim); #endif

1 0

master - str_list: also allow memory allocation without memory pool
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=097d14e64e845bdb… Commit: 097d14e64e845bdb1c0b8106e33f38f775e54fe5 Parent: 17196103e018a34eda016f6bdcd74395210cb114 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Tue Oct 20 14:09:09 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 15:47:55 2015 +0100 str_list: also allow memory allocation without memory pool --- lib/datastruct/str_list.c | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/lib/datastruct/str_list.c b/lib/datastruct/str_list.c index 1d3f08a..25ffddd 100644 --- a/lib/datastruct/str_list.c +++ b/lib/datastruct/str_list.c @@ -20,7 +20,8 @@ struct dm_list *str_list_create(struct dm_pool *mem) { struct dm_list *sl; - if (!(sl = dm_pool_alloc(mem, sizeof(struct dm_list)))) { + if (!(sl = mem ? dm_pool_alloc(mem, sizeof(struct dm_list)) + : dm_malloc(sizeof(struct dm_list)))) { log_errno(ENOMEM, "str_list allocation failed"); return NULL; } @@ -37,7 +38,8 @@ static int _str_list_add_no_dup_check(struct dm_pool *mem, struct dm_list *sll, if (!str) return_0; - if (!(sln = dm_pool_alloc(mem, sizeof(*sln)))) + if (!(sln = mem ? dm_pool_alloc(mem, sizeof(*sln)) + : dm_malloc(sizeof(*sln)))) return_0; sln->str = str;

1 0

master - New entries for pvs related fixes.
by David Teigland 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=17196103e018a34e… Commit: 17196103e018a34eda016f6bdcd74395210cb114 Parent: ccfc09f79b8d7ba15798a2b6f76545bb7fa3681e Author: David Teigland <teigland(a)redhat.com> AuthorDate: Fri Oct 30 09:05:21 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Fri Oct 30 09:05:21 2015 -0500 New entries for pvs related fixes. --- WHATS_NEW | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 00e807e..415efc7 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,7 @@ Version 2.02.133 - ====================================== + Fix missing PVs from pvs output if vgremove is run concurrently. + Remove unwanted error message when running pvs/vgs/lvs and vgremove at once. Check newly created VG's metadata do not overlap in metadata ring buffer. Check metadata area size is at least the minimum size defined for the format. Thin pool targets uses low_water_mark from profile.

1 0

master - metadata: format_text: also count with calculated mda size of 0
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=ccfc09f79b8d7ba1… Commit: ccfc09f79b8d7ba15798a2b6f76545bb7fa3681e Parent: 9a3b64e81ad1988f4143b5a4b4ac06eaf3035e58 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 30 12:02:29 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 12:02:34 2015 +0100 metadata: format_text: also count with calculated mda size of 0 When checking minimum mda size, make sure the mda_size after alignment and calculation is more than 0 - if there's no place for an MDA at the end of the disk, the _text_pv_add_metadata_area does not try to add it there and it returns (because we already have the MDA at the start of the disk at least). --- lib/format_text/format-text.c | 12 ++++++------ 1 files changed, 6 insertions(+), 6 deletions(-) diff --git a/lib/format_text/format-text.c b/lib/format_text/format-text.c index 29e9bd2..d4f2c89 100644 --- a/lib/format_text/format-text.c +++ b/lib/format_text/format-text.c @@ -2241,13 +2241,13 @@ static int _text_pv_add_metadata_area(const struct format_type *fmt, FMTu64 ").", pv_dev_name(pv), mda_size, limit_name, limit); - if (mda_size < MDA_SIZE_MIN) { - log_error("Metadata area size too small. " - "It must be at least %u bytes.", MDA_SIZE_MIN); - goto bad; - } - if (mda_size) { + if (mda_size < MDA_SIZE_MIN) { + log_error("Metadata area size too small: %" PRIu64" bytes. " + "It must be at least %u bytes.", mda_size, MDA_SIZE_MIN); + goto bad; + } + /* Wipe metadata area with zeroes. */ if (!dev_set((struct device *) pv->dev, mda_start, (size_t) ((mda_size > wipe_size) ?

1 0

master - tests: add test for minimum mda size
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=9a3b64e81ad1988f… Commit: 9a3b64e81ad1988f4143b5a4b4ac06eaf3035e58 Parent: c2e88d1107c496d2c1b39bbbfb9795e88b3a6b60 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 30 10:02:00 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 10:02:00 2015 +0100 tests: add test for minimum mda size --- test/shell/pvcreate-usage.sh | 9 +++++++++ 1 files changed, 9 insertions(+), 0 deletions(-) diff --git a/test/shell/pvcreate-usage.sh b/test/shell/pvcreate-usage.sh index 5b74ed3..c6aab26 100644 --- a/test/shell/pvcreate-usage.sh +++ b/test/shell/pvcreate-usage.sh @@ -12,6 +12,8 @@ test_description='Test pvcreate option values' SKIP_WITH_LVMPOLLD=1 PAGESIZE=$(getconf PAGESIZE) +# MDA_SIZE_MIN defined in lib/format_text/layout.h +MDA_SIZE_MIN=$((8*$PAGESIZE)) . lib/inittest @@ -23,6 +25,13 @@ not pvcreate --setphysicalvolumesize -1024 "$dev1" #COMM 'pvcreate rejects negative metadatasize' not pvcreate --metadatasize -1024 "$dev1" +#COMM 'pvcreate rejects metadatasize that is less than minimum size' +not pvcreate --dataalignment $((${MDA_SIZE_MIN}/2))b --metadatasize $((${MDA_SIZE_MIN}/2))b "$dev1" 2>err +grep "Metadata area size too small" err + +#COMM 'pvcreate accepts metadatasize that is at least the minimum size' +pvcreate --dataalignment ${MDA_SIZE_MIN}b --metadatasize ${MDA_SIZE_MIN}b "$dev1" + # x. metadatasize 0, defaults to 255 # FIXME: unable to check default value, not in reporting cmds # should default to 255 according to code

1 0

master - metadata: format_text: better check for metadata overlap
by Peter Rajnoha 30 Oct '15

30 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=c2e88d1107c496d2… Commit: c2e88d1107c496d2c1b39bbbfb9795e88b3a6b60 Parent: 406d8ff3323dde6860d88b603701aa4dace47af0 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 30 08:57:34 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 30 08:57:34 2015 +0100 metadata: format_text: better check for metadata overlap Actually, we don't need extra condition as introduced in commit 00348c0a630a7e5578edf48052ec695b15de752f. We should fix the last condition: (mdac->rlocn.size >= mdah->size) ...which should be: (MDA_HEADER_SIZE + (rlocn ? rlocn->size : 0) + mdac->rlocn.size >= mdah->size)) Where the "mdac" is new metadata, the "rlocn" is old metadata. So the main problem with the previous condition was that it didn't count in MDA_HEADER_SIZE properly (and possible existing metadata - the "rlocn"). This could have caused the error state where metadata in ring buffer overlap to not be hit. Replace the new condition introduced in 00348c0a630a7e5578edf48052ec695b15de752f with the improved one for the condition that existed there already but it was just incomplete. --- lib/format_text/format-text.c | 3 +-- 1 files changed, 1 insertions(+), 2 deletions(-) diff --git a/lib/format_text/format-text.c b/lib/format_text/format-text.c index 84d0947..29e9bd2 100644 --- a/lib/format_text/format-text.c +++ b/lib/format_text/format-text.c @@ -654,8 +654,7 @@ static int _vg_write_raw(struct format_instance *fid, struct volume_group *vg, if ((new_wrap && old_wrap) || (rlocn && (new_wrap || old_wrap) && (new_end > rlocn->offset)) || - (!rlocn && new_wrap && (new_end > mdac->rlocn.offset)) || - (mdac->rlocn.size >= mdah->size)) { + (MDA_HEADER_SIZE + (rlocn ? rlocn->size : 0) + mdac->rlocn.size >= mdah->size)) { log_error("VG %s metadata too large for circular buffer", vg->name); goto out;

1 0

master - WHATS_NEW: recent commits
by Peter Rajnoha 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=406d8ff3323dde68… Commit: 406d8ff3323dde6860d88b603701aa4dace47af0 Parent: 00348c0a630a7e5578edf48052ec695b15de752f Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Thu Oct 29 16:50:09 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Thu Oct 29 16:50:09 2015 +0100 WHATS_NEW: recent commits --- WHATS_NEW | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 03c70d6..00e807e 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,6 +1,7 @@ Version 2.02.133 - ====================================== - Check metadata area size is at least minimum size of 8 * memory page size. + Check newly created VG's metadata do not overlap in metadata ring buffer. + Check metadata area size is at least the minimum size defined for the format. Thin pool targets uses low_water_mark from profile. Dropping 'yet' from error of unsupported thick snapshot of snapshots. Do not support unpartitioned DASD devices with CDL formatted with pvcreate.

1 0

master - metadata: format_text: check VG metadata do not overlap themselves
by Peter Rajnoha 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=00348c0a630a7e55… Commit: 00348c0a630a7e5578edf48052ec695b15de752f Parent: ccb8da404d00288b7d49c7a28006ec5d4687bb55 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Thu Oct 29 16:33:06 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Thu Oct 29 16:46:41 2015 +0100 metadata: format_text: check VG metadata do not overlap themselves We're already checking whether old and new meta do not overlap in ring buffer (as we need to keep both old and new meta during vg_write up until vg_commit). We also need to check whether the new metadata do not overlap themselves in case we don't have old metadata yet (...because we're in vgcreate). This could happen if we're creating a VG so that the very first metadata written are long enough that it wraps themselves in metadata ring buffer. Although we limited the minimum metadata area size better with the previous commit ccb8da404d00288b7d49c7a28006ec5d4687bb55 which makes the initial VG metadata overlap in ring buffer to be less probable, the risk of hitting this overlap condition is still there if we still manage to generate big enough metadata somehow. For example, users can provide many and/or long VG tags during vgcreate so that the VG metadata is long enough to start to wrap in the ring buffer again... --- lib/format_text/format-text.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/lib/format_text/format-text.c b/lib/format_text/format-text.c index 6c653ca..84d0947 100644 --- a/lib/format_text/format-text.c +++ b/lib/format_text/format-text.c @@ -654,6 +654,7 @@ static int _vg_write_raw(struct format_instance *fid, struct volume_group *vg, if ((new_wrap && old_wrap) || (rlocn && (new_wrap || old_wrap) && (new_end > rlocn->offset)) || + (!rlocn && new_wrap && (new_end > mdac->rlocn.offset)) || (mdac->rlocn.size >= mdah->size)) { log_error("VG %s metadata too large for circular buffer", vg->name);

1 0

master - metadata: format_text: check metadata area size is at least MDA_SIZE_MIN
by Peter Rajnoha 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=ccb8da404d00288b… Commit: ccb8da404d00288b7d49c7a28006ec5d4687bb55 Parent: 28e54032c092ed4a5592575ac716fe698ce75d22 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Thu Oct 29 15:59:29 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Thu Oct 29 16:00:32 2015 +0100 metadata: format_text: check metadata area size is at least MDA_SIZE_MIN --- WHATS_NEW | 1 + lib/format_text/format-text.c | 7 ++++++- 2 files changed, 7 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index f453604..03c70d6 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Check metadata area size is at least minimum size of 8 * memory page size. Thin pool targets uses low_water_mark from profile. Dropping 'yet' from error of unsupported thick snapshot of snapshots. Do not support unpartitioned DASD devices with CDL formatted with pvcreate. diff --git a/lib/format_text/format-text.c b/lib/format_text/format-text.c index e0ec8ed..6c653ca 100644 --- a/lib/format_text/format-text.c +++ b/lib/format_text/format-text.c @@ -2146,7 +2146,6 @@ static int _text_pv_add_metadata_area(const struct format_type *fmt, goto bad; } /* Otherwise, give up and take any usable space. */ - /* FIXME: We should probably check for some minimum MDA size here. */ else mda_size = limit - mda_start; @@ -2242,6 +2241,12 @@ static int _text_pv_add_metadata_area(const struct format_type *fmt, FMTu64 ").", pv_dev_name(pv), mda_size, limit_name, limit); + if (mda_size < MDA_SIZE_MIN) { + log_error("Metadata area size too small. " + "It must be at least %u bytes.", MDA_SIZE_MIN); + goto bad; + } + if (mda_size) { /* Wipe metadata area with zeroes. */ if (!dev_set((struct device *) pv->dev, mda_start,

1 0

master - tests: update test for resize
by Zdenek Kabelac 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=28e54032c092ed4a… Commit: 28e54032c092ed4a5592575ac716fe698ce75d22 Parent: bca55c9b205ff2b0613ad44d85c12c6d6ae9691b Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 29 15:11:16 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 29 15:11:16 2015 +0100 tests: update test for resize Drop already tested 'threshold & create' which is in lvextend-thin-full.sh Count with now match faster 'dmeventd' wakeup on watermark as it's now nearly instant after crossing threshold value. --- test/shell/lvextend-thin-metadata-dmeventd.sh | 28 +++++------------------- 1 files changed, 6 insertions(+), 22 deletions(-) diff --git a/test/shell/lvextend-thin-metadata-dmeventd.sh b/test/shell/lvextend-thin-metadata-dmeventd.sh index 030991f..a2d6a39 100644 --- a/test/shell/lvextend-thin-metadata-dmeventd.sh +++ b/test/shell/lvextend-thin-metadata-dmeventd.sh @@ -23,8 +23,8 @@ meta_percent_() { wait_for_change_() { # dmeventd only checks every 10 seconds :( - for i in $(seq 1 15) ; do - test "$(meta_percent_)" != "$1" && return + for i in $(seq 1 12) ; do + test "$(meta_percent_)" -lt "$1" && return sleep 1 done @@ -101,43 +101,27 @@ lvconvert -y --chunksize 64k --thinpool $vg/pool --poolmetadata $vg/$lv1 vgchange -ay $vg -# Check dmeventd resizes metadata +# Check dmeventd resizes metadata via timeout (nothing is written to pool) pre=$(meta_percent_) wait_for_change_ $pre lvchange -an $vg -lvs -a $vg # fake_metadata_ 350 2 >data lvchange -ay $vg/$lv1 "$LVM_TEST_THIN_RESTORE_CMD" -i data -o "$DM_DEV_DIR/mapper/$vg-$lv1" -lvs -a $vg -dmsetup table lvconvert -y --chunksize 64k --thinpool $vg/pool --poolmetadata $vg/$lv1 lvchange -ay $vg/pool $vg/$lv1 lvs -a $vg lvcreate -s -Ky -n $lv2 $vg/thin -echo 2 >"$DM_DEV_DIR/mapper/$vg-$lv2" - -#lvchange -an $vg -#lvconvert -y --chunksize 64k --thinpool $vg/pool --poolmetadata $vg/$lv1 -#lvchange -ay $vg/$lv1 -#thin_dump "$DM_DEV_DIR/mapper/$vg-$lv1" -#exit +pre=$(meta_percent_) -# no more space for new thin LV -# -# TODO: -# though maybe 'lvcreate' itself should initiate resize - if dmeventd is not 'fast' enough -# deploying usage of threshold kernel module paramater would likely help as well -# for now it stops here: -not lvcreate -s -Ky -n $lv3 $vg/thin -lvs -a $vg +# go over thin metadata threshold +echo 2 >"$DM_DEV_DIR/mapper/$vg-$lv2" -pre=$(meta_percent_) wait_for_change_ $pre lvs -a $vg

1 0

master - tests: replace invalid use of 'fail' with 'die'
by okozina 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=bca55c9b205ff2b0… Commit: bca55c9b205ff2b0613ad44d85c12c6d6ae9691b Parent: f104a81932e1011acb4798402f4f67f78f701a0b Author: Ondrej Kozina <okozina(a)redhat.com> AuthorDate: Thu Oct 29 13:30:29 2015 +0100 Committer: Ondrej Kozina <okozina(a)redhat.com> CommitterDate: Thu Oct 29 13:30:29 2015 +0100 tests: replace invalid use of 'fail' with 'die' --- test/shell/lvconvert-mirror.sh | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/test/shell/lvconvert-mirror.sh b/test/shell/lvconvert-mirror.sh index 029112c..5c7aad3 100644 --- a/test/shell/lvconvert-mirror.sh +++ b/test/shell/lvconvert-mirror.sh @@ -317,7 +317,7 @@ lvremove -ff $vg # to get synced lvcreate -l2 -n $lv1 $vg lvconvert --type mirror -i1 -m1 $vg/$lv1 | tee out -grep -e "$vg/$lv1: Converted:" out || fail "Missing sync info in foreground mode" +grep -e "$vg/$lv1: Converted:" out || die "Missing sync info in foreground mode" lvremove -ff $vg vgremove -ff $vg

1 0

master - tests: let pass bigger readahead
by Zdenek Kabelac 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=f104a81932e1011a… Commit: f104a81932e1011acb4798402f4f67f78f701a0b Parent: 3720eb63bed5d087a13d184894197373ebbbaeba Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 29 12:33:15 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 29 12:39:07 2015 +0100 tests: let pass bigger readahead If the underlaying device has actually bigger read-ahead settings, let it pass. But anyway switch to 512 strip-size to get really high R-A sector count. --- test/shell/read-ahead.sh | 5 ++--- 1 files changed, 2 insertions(+), 3 deletions(-) diff --git a/test/shell/read-ahead.sh b/test/shell/read-ahead.sh index 82a0707..94465cb 100644 --- a/test/shell/read-ahead.sh +++ b/test/shell/read-ahead.sh @@ -21,12 +21,11 @@ SKIP_WITH_LVMPOLLD=1 aux prepare_vg 5 #COMM "test various read ahead settings (bz450922)" -lvcreate -l 100%FREE -i5 -I256 -n $lv $vg +lvcreate -l 100%FREE -i5 -I512 -n $lv $vg ra=$(get lv_field $vg/$lv lv_kernel_read_ahead --units s --nosuffix) -test $(( ( $ra / 5 ) * 5 )) -eq $ra +test $(( ( $ra / 5 ) * 5 )) -le $ra not lvchange -r auto $vg/$lv 2>&1 | grep auto check lv_field $vg/$lv lv_read_ahead auto -check lv_field $vg/$lv lv_kernel_read_ahead 5120 --units s --nosuffix lvchange -r 640 $vg/$lv check lv_field $vg/$lv lv_read_ahead 640 --units s --nosuffix lvremove -ff $vg

1 0

master - tests: fix wrong line has been commented
by Zdenek Kabelac 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=3720eb63bed5d087… Commit: 3720eb63bed5d087a13d184894197373ebbbaeba Parent: 8b5525383f2070a6df8508bfe5696324ca6acb54 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 29 12:30:00 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 29 12:39:07 2015 +0100 tests: fix wrong line has been commented --- test/shell/snapshot-remove-dmsetup.sh | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/test/shell/snapshot-remove-dmsetup.sh b/test/shell/snapshot-remove-dmsetup.sh index 395fad1..6a4f9d2 100644 --- a/test/shell/snapshot-remove-dmsetup.sh +++ b/test/shell/snapshot-remove-dmsetup.sh @@ -36,12 +36,12 @@ mount -o errors=remount-ro "$DM_DEV_DIR/$vg1/snap" mnt sync -# intetionally suspend layer bellow +# intentionally suspend layer bellow dmsetup suspend $vg-$lv1 # now this should pass without blocking -#dmsetup suspend --noflush --nolockfs $vg1-snap & -dmsetup suspend $vg1-snap & +dmsetup suspend --noflush --nolockfs $vg1-snap & +#dmsetup suspend $vg1-snap & sleep .5

1 0

master - tests: no point in using should
by Zdenek Kabelac 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=8b5525383f2070a6… Commit: 8b5525383f2070a6df8508bfe5696324ca6acb54 Parent: f58c63410327c2bc49b671df4538dee9e90e577d Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 29 12:23:38 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 29 12:39:07 2015 +0100 tests: no point in using should lvmetad does not support lvm1 - so expect failure. --- test/shell/lvmetad-lvm1.sh | 15 ++++++++++----- 1 files changed, 10 insertions(+), 5 deletions(-) diff --git a/test/shell/lvmetad-lvm1.sh b/test/shell/lvmetad-lvm1.sh index 3b14fdd..c1b34ea 100644 --- a/test/shell/lvmetad-lvm1.sh +++ b/test/shell/lvmetad-lvm1.sh @@ -14,14 +14,19 @@ SKIP_WITH_LVMPOLLD=1 . lib/inittest +# lvmetad does not support lvm1 format + aux prepare_devs 2 pvcreate --metadatatype 1 "$dev1" -should vgscan --cache -pvs | should grep "$dev1" +not vgscan --cache +pvs | tee out +not grep "$dev1" out vgcreate --metadatatype 1 $vg1 "$dev1" -should vgscan --cache -vgs | should grep $vg1 -pvs | should grep "$dev1" +not vgscan --cache +vgs | tee out +not grep $vg1 out +pvs | tee out +not grep "$dev1" out # check for RHBZ 1080189 -- SEGV in lvremove/vgremove pvcreate -ff -y --metadatatype 1 "$dev1" "$dev2"

1 0

master - cleanup: error is not a WARNING
by Zdenek Kabelac 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=f58c63410327c2bc… Commit: f58c63410327c2bc49b671df4538dee9e90e577d Parent: 175119fdcd1a8eb23011e4c25d0df5899dfc42ea Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 29 12:18:40 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 29 12:38:59 2015 +0100 cleanup: error is not a WARNING Drop 'WARNING' from error message. It's plain error message leading to command failure. --- lib/cache/lvmetad.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/lib/cache/lvmetad.c b/lib/cache/lvmetad.c index c7f6ecb..683c87c 100644 --- a/lib/cache/lvmetad.c +++ b/lib/cache/lvmetad.c @@ -1331,7 +1331,7 @@ int lvmetad_pvscan_single(struct cmd_context *cmd, struct device *dev, log_warn("WARNING: Ignoring obsolete format of metadata (%s) on device %s when using lvmetad", baton.fid->fmt->name, dev_name(dev)); else - log_error("WARNING: Ignoring obsolete format of metadata (%s) on device %s when using lvmetad", + log_error("Ignoring obsolete format of metadata (%s) on device %s when using lvmetad.", baton.fid->fmt->name, dev_name(dev)); lvmcache_fmt(info)->ops->destroy_instance(baton.fid);

1 0

master - cleanup: remove thin low_water_mark from metadata
by Zdenek Kabelac 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=175119fdcd1a8eb2… Commit: 175119fdcd1a8eb23011e4c25d0df5899dfc42ea Parent: 33a8a2febf483df088ec8e1928d25a6705a18991 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 29 12:07:24 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 29 12:14:20 2015 +0100 cleanup: remove thin low_water_mark from metadata This option could never have been printed in lvm2 metadata, so it could be safely removed as it could have been set only as 0. These configurable setting is supported via metadata profile. --- lib/metadata/lv_manip.c | 2 -- lib/metadata/metadata-exported.h | 1 - lib/thin/thin.c | 7 ------- tools/lvconvert.c | 1 - 4 files changed, 0 insertions(+), 11 deletions(-) diff --git a/lib/metadata/lv_manip.c b/lib/metadata/lv_manip.c index 8e20dfa..2310bfa 100644 --- a/lib/metadata/lv_manip.c +++ b/lib/metadata/lv_manip.c @@ -7388,8 +7388,6 @@ static struct logical_volume *_lv_create_an_lv(struct volume_group *vg, first_seg(lv)->chunk_size = lp->chunk_size; first_seg(lv)->zero_new_blocks = lp->zero ? 1 : 0; first_seg(lv)->discards = lp->discards; - /* FIXME: use lowwatermark via lvm.conf global for all thinpools ? */ - first_seg(lv)->low_water_mark = 0; if (!recalculate_pool_chunk_size_with_dev_hints(lv, lp->passed_args, lp->thin_chunk_size_calc_policy)) { stack; diff --git a/lib/metadata/metadata-exported.h b/lib/metadata/metadata-exported.h index 41838a8..2d118f4 100644 --- a/lib/metadata/metadata-exported.h +++ b/lib/metadata/metadata-exported.h @@ -454,7 +454,6 @@ struct lv_segment { struct lv_segment_area *meta_areas; /* For RAID */ struct logical_volume *metadata_lv; /* For thin_pool */ uint64_t transaction_id; /* For thin_pool, thin */ - uint64_t low_water_mark; /* For thin_pool */ unsigned zero_new_blocks; /* For thin_pool */ thin_discards_t discards; /* For thin_pool */ struct dm_list thin_messages; /* For thin_pool */ diff --git a/lib/thin/thin.c b/lib/thin/thin.c index ad400ef..be7a6b4 100644 --- a/lib/thin/thin.c +++ b/lib/thin/thin.c @@ -118,10 +118,6 @@ static int _thin_pool_text_import(struct lv_segment *seg, else if (!set_pool_discards(&seg->discards, discards_str)) return SEG_LOG_ERROR("Discards option unsupported for"); - if (dm_config_has_node(sn, "low_water_mark") && - !dm_config_get_uint64(sn, "low_water_mark", &seg->low_water_mark)) - return SEG_LOG_ERROR("Could not read low_water_mark"); - if ((seg->chunk_size < DM_THIN_MIN_DATA_BLOCK_SIZE) || (seg->chunk_size > DM_THIN_MAX_DATA_BLOCK_SIZE)) return SEG_LOG_ERROR("Unsupported value %u for chunk_size", @@ -169,9 +165,6 @@ static int _thin_pool_text_export(const struct lv_segment *seg, struct formatter return 0; } - if (seg->low_water_mark) - outf(f, "low_water_mark = %" PRIu64, seg->low_water_mark); - if (seg->zero_new_blocks) outf(f, "zero_new_blocks = 1"); diff --git a/tools/lvconvert.c b/tools/lvconvert.c index 0e06160..a7df26d 100644 --- a/tools/lvconvert.c +++ b/tools/lvconvert.c @@ -3064,7 +3064,6 @@ static int _lvconvert_pool(struct cmd_context *cmd, /* FIXME: revert renamed LVs in fail path? */ /* FIXME: any common code with metadata/thin_manip.c extend_pool() ? */ - seg->low_water_mark = 0; seg->transaction_id = 0; mda_write:

1 0

master - cleanup: use same type
by Zdenek Kabelac 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=33a8a2febf483df0… Commit: 33a8a2febf483df088ec8e1928d25a6705a18991 Parent: f32f0bd2a7aeddb90d41f342d4213ef054e9d4fd Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 29 12:01:31 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 29 12:14:20 2015 +0100 cleanup: use same type --- lib/metadata/thin_manip.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/lib/metadata/thin_manip.c b/lib/metadata/thin_manip.c index 6b22ce5..2d3146f 100644 --- a/lib/metadata/thin_manip.c +++ b/lib/metadata/thin_manip.c @@ -216,7 +216,7 @@ int thin_pool_feature_supported(const struct logical_volume *lv, int feature) int pool_below_threshold(const struct lv_segment *pool_seg) { dm_percent_t percent; - int threshold = DM_PERCENT_1 * + dm_percent_t threshold = DM_PERCENT_1 * find_config_tree_int(pool_seg->lv->vg->cmd, activation_thin_pool_autoextend_threshold_CFG, lv_config_profile(pool_seg->lv));

1 0

master - cleanup: easier error messages
by Zdenek Kabelac 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=f32f0bd2a7aeddb9… Commit: f32f0bd2a7aeddb90d41f342d4213ef054e9d4fd Parent: 99237f0908d87592815f4bdf3c239e8a108e835c Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 29 11:01:24 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 29 12:14:20 2015 +0100 cleanup: easier error messages --- lib/thin/thin.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/lib/thin/thin.c b/lib/thin/thin.c index d042473..ad400ef 100644 --- a/lib/thin/thin.c +++ b/lib/thin/thin.c @@ -279,20 +279,20 @@ static int _thin_pool_add_target_line(struct dev_manager *dm, if (!(attr & THIN_FEATURE_BLOCK_SIZE) && (seg->chunk_size & (seg->chunk_size - 1))) { - log_error("Thin pool target does not support %uKiB chunk size " - "(needs kernel >= 3.6).", seg->chunk_size / 2); + log_error("Thin pool target does not support %s chunk size (needs" + " kernel >= 3.6).", display_size(cmd, seg->chunk_size)); return 0; } if (!(metadata_dlid = build_dm_uuid(mem, seg->metadata_lv, NULL))) { log_error("Failed to build uuid for metadata LV %s.", - seg->metadata_lv->name); + display_lvname(seg->metadata_lv)); return 0; } if (!(pool_dlid = build_dm_uuid(mem, seg_lv(seg, 0), NULL))) { log_error("Failed to build uuid for pool LV %s.", - seg_lv(seg, 0)->name); + display_lvname(seg_lv(seg, 0))); return 0; }

1 0

master - thin: enable usage of kernel low_water_mark
by Zdenek Kabelac 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=99237f0908d87592… Commit: 99237f0908d87592815f4bdf3c239e8a108e835c Parent: 099466939e4d6bdbe4b105f6b052dee6fc99b9d4 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 29 11:58:06 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 29 12:14:20 2015 +0100 thin: enable usage of kernel low_water_mark Now with correctly functioning dmeventd enable usage of low_water_mark for faster reaction on pool's threshold. When user select e.g. 80% as a threshold value, dmeventd doesn't need to wait 10 seconds till monitoring timer expires, but nearly instantly resizes thin-pool to fit bellow threshold. --- WHATS_NEW | 1 + lib/thin/thin.c | 14 +++++++++++++- 2 files changed, 14 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 869e506..f453604 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Thin pool targets uses low_water_mark from profile. Dropping 'yet' from error of unsupported thick snapshot of snapshots. Do not support unpartitioned DASD devices with CDL formatted with pvcreate. For thins use flush for suspend only when volume size is reduced. diff --git a/lib/thin/thin.c b/lib/thin/thin.c index 7bf7f4e..d042473 100644 --- a/lib/thin/thin.c +++ b/lib/thin/thin.c @@ -266,6 +266,8 @@ static int _thin_pool_add_target_line(struct dev_manager *dm, struct lvinfo info; uint64_t transaction_id = 0; unsigned attr; + uint64_t low_water_mark; + int threshold; if (!_thin_target_present(cmd, NULL, &attr)) return_0; @@ -294,10 +296,20 @@ static int _thin_pool_add_target_line(struct dev_manager *dm, return 0; } + threshold = find_config_tree_int(seg->lv->vg->cmd, + activation_thin_pool_autoextend_threshold_CFG, + lv_config_profile(seg->lv)); + if (threshold < 50) + threshold = 50; + if (threshold < 100) + low_water_mark = (len * threshold + 99) / 100; + else + low_water_mark = len; + if (!dm_tree_node_add_thin_pool_target(node, len, seg->transaction_id, metadata_dlid, pool_dlid, - seg->chunk_size, seg->low_water_mark, + seg->chunk_size, low_water_mark, seg->zero_new_blocks ? 0 : 1)) return_0;

1 0

master - thin: dmeventd plugin check number of failures
by Zdenek Kabelac 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=099466939e4d6bdb… Commit: 099466939e4d6bdbe4b105f6b052dee6fc99b9d4 Parent: b3c81d02c974eda1c33afd88517d899e64aea24e Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 29 11:52:11 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 29 12:14:20 2015 +0100 thin: dmeventd plugin check number of failures If plugin's lvm command execution fails too often (>10 times), there is no point to torture system more then necessary, just log and drop monitoring in this case. --- WHATS_NEW_DM | 1 + daemons/dmeventd/plugins/thin/dmeventd_thin.c | 19 ++++++++++++++++--- 2 files changed, 17 insertions(+), 3 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 961aec5..63cb2cc 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Disable thin monitoring plugin when it fails too often (>10 times). Fix/restore parsing of empty field '-' when processing dmeventd event. Enhance dm_tree_node_size_changed() to recognize size reduction. Support exit on idle for dmenventd (1 hour). diff --git a/daemons/dmeventd/plugins/thin/dmeventd_thin.c b/daemons/dmeventd/plugins/thin/dmeventd_thin.c index 888cc0c..7f422a9 100644 --- a/daemons/dmeventd/plugins/thin/dmeventd_thin.c +++ b/daemons/dmeventd/plugins/thin/dmeventd_thin.c @@ -18,6 +18,7 @@ #include <sys/wait.h> #include <stdarg.h> +#include <pthread.h> /* TODO - move this mountinfo code into library to be reusable */ #ifdef __linux__ @@ -36,6 +37,8 @@ #define UMOUNT_COMMAND "/bin/umount" +#define MAX_FAILS (10) + #define THIN_DEBUG 0 struct dso_state { @@ -44,6 +47,7 @@ struct dso_state { int data_percent_check; uint64_t known_metadata_size; uint64_t known_data_size; + unsigned fails; char cmd_str[1024]; }; @@ -157,7 +161,7 @@ static int _run(const char *cmd, ...) argv = alloca(sizeof(const char *) * (argc + 1)); argv[0] = cmd; - va_start(ap, cmd); + va_start(ap, cmd); while ((argv[++i] = va_arg(ap, const char *))); va_end(ap); @@ -245,7 +249,9 @@ static void _use_policy(struct dm_task *dmt, struct dso_state *state) log_error("Failed to extend thin pool %s.", dm_task_get_name(dmt)); _umount(dmt); - } + state->fails++; + } else + state->fails = 0; } void process_event(struct dm_task *dmt, @@ -270,7 +276,7 @@ void process_event(struct dm_task *dmt, if (event & DM_EVENT_DEVICE_ERROR) { /* Error -> no need to check and do instant resize */ _use_policy(dmt, state); - return; + goto out; } dm_get_next_target(dmt, next, &start, &length, &target_type, &params); @@ -338,6 +344,13 @@ void process_event(struct dm_task *dmt, out: if (tps) dm_pool_free(state->mem, tps); + + if (state->fails >= MAX_FAILS) { + log_warn("WARNING: Dropping monitoring of %s. " + "lvm2 command fails too often (%u times in raw).", + device, state->fails); + pthread_kill(pthread_self(), SIGALRM); + } } int register_device(const char *device,

1 0

master - revert: 3d03e504cda6307f692c5879f6850b2e27af7d71: message about VG metadata size vs. PV mda size
by Peter Rajnoha 29 Oct '15

29 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=b3c81d02c974eda1… Commit: b3c81d02c974eda1c33afd88517d899e64aea24e Parent: 5886ff64ebcac8c1b53c665c598728e481590964 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Thu Oct 29 11:08:28 2015 +0100 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Thu Oct 29 11:10:48 2015 +0100 revert: 3d03e504cda6307f692c5879f6850b2e27af7d71: message about VG metadata size vs. PV mda size The message needs refinement - it's not correct in all situations. --- lib/format_text/format-text.c | 7 ++----- 1 files changed, 2 insertions(+), 5 deletions(-) diff --git a/lib/format_text/format-text.c b/lib/format_text/format-text.c index c1c4787..e0ec8ed 100644 --- a/lib/format_text/format-text.c +++ b/lib/format_text/format-text.c @@ -655,11 +655,8 @@ static int _vg_write_raw(struct format_instance *fid, struct volume_group *vg, if ((new_wrap && old_wrap) || (rlocn && (new_wrap || old_wrap) && (new_end > rlocn->offset)) || (mdac->rlocn.size >= mdah->size)) { - log_error("VG %s metadata too large: size of metadata to write " - "is %" PRIu64 " bytes while PV metadata area size " - "on %s is %" PRIu64 " bytes.", - vg->name, mdac->rlocn.size, - dev_name(mdac->area.dev), mdah->size); + log_error("VG %s metadata too large for circular buffer", + vg->name); goto out; }

1 0

master - pvs: don't treat duplicate PVs as missed
by David Teigland 27 Oct '15

27 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=5886ff64ebcac8c1… Commit: 5886ff64ebcac8c1b53c665c598728e481590964 Parent: a4418b34c150299aa88c6540bb4e3680f8bb4dae Author: David Teigland <teigland(a)redhat.com> AuthorDate: Tue Oct 27 12:03:57 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Tue Oct 27 12:03:57 2015 -0500 pvs: don't treat duplicate PVs as missed The recent addition to check for PVs that were missed during the first iteration of processing was unintentionally catching duplicate PVs because duplicates were not removed from the all_devices list when the primary dev was processed. Also change a message from warn back to verbose. --- tools/toollib.c | 24 +++++++++++++++++++++--- 1 files changed, 21 insertions(+), 3 deletions(-) diff --git a/tools/toollib.c b/tools/toollib.c index 3804d66..42ac71a 100644 --- a/tools/toollib.c +++ b/tools/toollib.c @@ -2885,12 +2885,17 @@ static int _process_pvs_in_vg(struct cmd_context *cmd, } /* + * We have processed the PV on device pv->dev. Now + * deal with any duplicates of this PV on other + * devices. + */ + + /* * This is a very rare and obscure case where multiple * duplicate devices are specified on the command line * referring to this PV. In this case we want to * process this PV once for each specified device. */ - if (!skip && !dm_list_empty(arg_devices)) { while ((dil = _device_list_find_pvid(arg_devices, pv))) { _device_list_remove(arg_devices, dil->dev); @@ -2925,7 +2930,6 @@ static int _process_pvs_in_vg(struct cmd_context *cmd, * we want each of them to be displayed in the context * of this VG, so that this VG name appears next to it. */ - if (process_all_devices && lvmcache_found_duplicate_pvs()) { while ((dil = _device_list_find_pvid(all_devices, pv))) { _device_list_remove(all_devices, dil->dev); @@ -2942,6 +2946,20 @@ static int _process_pvs_in_vg(struct cmd_context *cmd, lvmcache_replace_dev(cmd, pv, dev_orig); } } + + /* + * Remove any duplicates of the processed device from + * the list of all devices. If they were left in the + * list of all devices, they would be considered + * "missed" at the end. + */ + if (process_all_pvs && lvmcache_found_duplicate_pvs()) { + while ((dil = _device_list_find_pvid(all_devices, pv))) { + log_very_verbose("Skip duplicate device %s of processed device %s", + dev_name(dil->dev), dev_name(pv->dev)); + _device_list_remove(all_devices, dil->dev); + } + } } /* @@ -3160,7 +3178,7 @@ int process_each_pv(struct cmd_context *cmd, dm_list_init(&arg_missed_orig); _device_list_copy(cmd, &arg_missed, &arg_missed_orig); - log_warn("Some PVs were not found in first search, retrying."); + log_verbose("Some PVs were not found in first search, retrying."); lvmcache_destroy(cmd, 0, 0); lvmcache_init();

1 0

master - vgs, lvs: ignore error if VG is removed
by David Teigland 27 Oct '15

27 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=a4418b34c150299a… Commit: a4418b34c150299aa88c6540bb4e3680f8bb4dae Parent: 65ec00ce202411f6e6f91955ab9422d44bb86a9b Author: David Teigland <teigland(a)redhat.com> AuthorDate: Tue Oct 27 10:52:01 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Tue Oct 27 10:52:01 2015 -0500 vgs, lvs: ignore error if VG is removed If a VG is removed between the time that 'vgs' or 'lvs' (with no args) creates the list of VGs and the time that it reads the VG to process it, then ignore the removed VG; don't report an error that it could not be found, since it wasn't named by the command. --- tools/toollib.c | 26 ++++++++++++++++---------- 1 files changed, 16 insertions(+), 10 deletions(-) diff --git a/tools/toollib.c b/tools/toollib.c index 59ba29f..3804d66 100644 --- a/tools/toollib.c +++ b/tools/toollib.c @@ -1909,7 +1909,7 @@ int select_match_pv(struct cmd_context *cmd, struct processing_handle *handle, return 1; } -static int _process_vgnameid_list(struct cmd_context *cmd, uint32_t flags, +static int _process_vgnameid_list(struct cmd_context *cmd, uint32_t read_flags, struct dm_list *vgnameids_to_process, struct dm_list *arg_vgnames, struct dm_list *arg_tags, @@ -1952,8 +1952,8 @@ static int _process_vgnameid_list(struct cmd_context *cmd, uint32_t flags, continue; } - vg = vg_read(cmd, vg_name, vg_uuid, flags, lockd_state); - if (_ignore_vg(vg, vg_name, arg_vgnames, flags, &skip, &notfound)) { + vg = vg_read(cmd, vg_name, vg_uuid, read_flags, lockd_state); + if (_ignore_vg(vg, vg_name, arg_vgnames, read_flags, &skip, &notfound)) { stack; ret_max = ECMD_FAILED; goto endvg; @@ -2020,7 +2020,7 @@ static int _copy_str_to_vgnameid_list(struct cmd_context *cmd, struct dm_list *s * Call process_single_vg() for each VG selected by the command line arguments. */ int process_each_vg(struct cmd_context *cmd, int argc, char **argv, - uint32_t flags, struct processing_handle *handle, + uint32_t read_flags, struct processing_handle *handle, process_single_vg_fn_t process_single_vg) { int handle_supplied = handle != NULL; @@ -2070,6 +2070,9 @@ int process_each_vg(struct cmd_context *cmd, int argc, char **argv, goto out; } + if (dm_list_empty(&arg_vgnames)) + read_flags |= READ_OK_NOTFOUND; + /* * If we obtained a full list of VGs on the system, we need to work through them all; * otherwise we can merely work through the VG names provided. @@ -2086,7 +2089,7 @@ int process_each_vg(struct cmd_context *cmd, int argc, char **argv, !init_selection_handle(cmd, handle, VGS)) goto_out; - ret = _process_vgnameid_list(cmd, flags, &vgnameids_to_process, + ret = _process_vgnameid_list(cmd, read_flags, &vgnameids_to_process, &arg_vgnames, &arg_tags, handle, process_single_vg); out: if (!handle_supplied) @@ -2375,7 +2378,7 @@ static int _get_arg_lvnames(struct cmd_context *cmd, return ret_max; } -static int _process_lv_vgnameid_list(struct cmd_context *cmd, uint32_t flags, +static int _process_lv_vgnameid_list(struct cmd_context *cmd, uint32_t read_flags, struct dm_list *vgnameids_to_process, struct dm_list *arg_vgnames, struct dm_list *arg_lvnames, @@ -2442,8 +2445,8 @@ static int _process_lv_vgnameid_list(struct cmd_context *cmd, uint32_t flags, continue; } - vg = vg_read(cmd, vg_name, vg_uuid, flags, lockd_state); - if (_ignore_vg(vg, vg_name, arg_vgnames, flags, &skip, &notfound)) { + vg = vg_read(cmd, vg_name, vg_uuid, read_flags, lockd_state); + if (_ignore_vg(vg, vg_name, arg_vgnames, read_flags, &skip, &notfound)) { stack; ret_max = ECMD_FAILED; goto endvg; @@ -2471,7 +2474,7 @@ endvg: /* * Call process_single_lv() for each LV selected by the command line arguments. */ -int process_each_lv(struct cmd_context *cmd, int argc, char **argv, uint32_t flags, +int process_each_lv(struct cmd_context *cmd, int argc, char **argv, uint32_t read_flags, struct processing_handle *handle, process_single_lv_fn_t process_single_lv) { int handle_supplied = handle != NULL; @@ -2538,6 +2541,9 @@ int process_each_lv(struct cmd_context *cmd, int argc, char **argv, uint32_t fla goto out; } + if (dm_list_empty(&arg_vgnames)) + read_flags |= READ_OK_NOTFOUND; + /* * If we obtained a full list of VGs on the system, we need to work through them all; * otherwise we can merely work through the VG names provided. @@ -2547,7 +2553,7 @@ int process_each_lv(struct cmd_context *cmd, int argc, char **argv, uint32_t fla else if ((ret = _copy_str_to_vgnameid_list(cmd, &arg_vgnames, &vgnameids_to_process)) != ECMD_PROCESSED) goto_out; - ret = _process_lv_vgnameid_list(cmd, flags, &vgnameids_to_process, &arg_vgnames, &arg_lvnames, + ret = _process_lv_vgnameid_list(cmd, read_flags, &vgnameids_to_process, &arg_vgnames, &arg_lvnames, &arg_tags, handle, process_single_lv); out: if (!handle_supplied)

1 0

master - device: Tidy DASD CDL format detection code.
by Alasdair Kergon 27 Oct '15

27 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=65ec00ce202411f6… Commit: 65ec00ce202411f6e6f91955ab9422d44bb86a9b Parent: 6e1e0e881368be6337599dcf7f627f785ebf415c Author: Alasdair G Kergon <agk(a)redhat.com> AuthorDate: Tue Oct 27 15:27:52 2015 +0000 Committer: Alasdair G Kergon <agk(a)redhat.com> CommitterDate: Tue Oct 27 15:27:52 2015 +0000 device: Tidy DASD CDL format detection code. --- lib/device/dev-dasd.c | 85 +++++++++++++++++++++++++++++++++++-------------- 1 files changed, 61 insertions(+), 24 deletions(-) diff --git a/lib/device/dev-dasd.c b/lib/device/dev-dasd.c index 28156ca..220293e 100644 --- a/lib/device/dev-dasd.c +++ b/lib/device/dev-dasd.c @@ -17,7 +17,28 @@ #include "dev-type.h" #include <sys/ioctl.h> -typedef struct dasd_information_t { +#ifdef __linux__ + +/* + * Interface taken from kernel header arch/s390/include/uapi/asm/dasd.h + */ + +/* + * Author(s)......: Holger Smolinski <Holger.Smolinski(a)de.ibm.com> + * Copyright IBM Corp. 1999, 2000 + * EMC Symmetrix ioctl Copyright EMC Corporation, 2008 + * Author.........: Nigel Hislop <hislop_nigel(a)emc.com> + */ + +#define DASD_IOCTL_LETTER 'D' +#define DASD_API_VERSION 6 + +/* + * struct dasd_information2_t + * represents any data about the device, which is visible to userspace. + * including foramt and featueres. + */ +typedef struct dasd_information2_t { unsigned int devno; /* S/390 devno */ unsigned int real_devno; /* for aliases */ unsigned int schid; /* S/390 subchannel identifier */ @@ -34,41 +55,57 @@ typedef struct dasd_information_t { unsigned int FBA_layout; /* fixed block size (like AIXVOL) */ unsigned int characteristics_size; unsigned int confdata_size; - unsigned char characteristics[64];/*from read_device_characteristics */ - unsigned char configuration_data[256];/*from read_configuration_data */ + char characteristics[64]; /* from read_device_characteristics */ + char configuration_data[256]; /* from read_configuration_data */ unsigned int format; /* format info like formatted/cdl/ldl/... */ - unsigned int features; /* dasd features like 'ro',... */ - unsigned int reserved0; /* reserved for further use ,... */ - unsigned int reserved1; /* reserved for further use ,... */ - unsigned int reserved2; /* reserved for further use ,... */ - unsigned int reserved3; /* reserved for further use ,... */ - unsigned int reserved4; /* reserved for further use ,... */ - unsigned int reserved5; /* reserved for further use ,... */ - unsigned int reserved6; /* reserved for further use ,... */ - unsigned int reserved7; /* reserved for further use ,... */ -} dasd_information_t; - -#define DASD_FORMAT_CDL 2 -#define BIODASDINFO2 _IOR('D', 3, dasd_information_t) + unsigned int features; /* dasd features like 'ro',... */ + unsigned int reserved0; /* reserved for further use ,... */ + unsigned int reserved1; /* reserved for further use ,... */ + unsigned int reserved2; /* reserved for further use ,... */ + unsigned int reserved3; /* reserved for further use ,... */ + unsigned int reserved4; /* reserved for further use ,... */ + unsigned int reserved5; /* reserved for further use ,... */ + unsigned int reserved6; /* reserved for further use ,... */ + unsigned int reserved7; /* reserved for further use ,... */ +} dasd_information2_t; + +#define DASD_FORMAT_CDL 2 + +/* Get information on a dasd device (enhanced) */ +#define BIODASDINFO2 _IOR(DASD_IOCTL_LETTER,3,dasd_information2_t) + +/* + * End of included interface. + */ int dasd_is_cdl_formatted(struct device *dev) { int ret = 0; - dasd_information_t dasd_info; + dasd_information2_t dasd_info2; - if (!dev_open_readonly(dev)) { - stack; - return ret; - } + if (!dev_open_readonly(dev)) + return_0; - if (ioctl(dev->fd, BIODASDINFO2, &dasd_info) != 0) - goto_out; + if (ioctl(dev->fd, BIODASDINFO2, &dasd_info2)) { + log_sys_error("ioctl BIODASDINFO2", dev_name(dev)); + goto out; + } - if (dasd_info.format == DASD_FORMAT_CDL) + if (dasd_info2.format == DASD_FORMAT_CDL) ret = 1; + out: if (!dev_close(dev)) stack; return ret; } + +#else + +int dasd_is_cdl_formatted(struct device *dev) +{ + return 0; +} + +#endif

1 0

master - tests: using matching type
by Zdenek Kabelac 27 Oct '15

27 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=6e1e0e881368be63… Commit: 6e1e0e881368be6337599dcf7f627f785ebf415c Parent: 4159680a0ee7eceb4d5124c1e4e78b418b12f357 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 27 15:33:59 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 27 16:00:10 2015 +0100 tests: using matching type Compare time_t. --- test/lib/brick-shelltest.h | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/test/lib/brick-shelltest.h b/test/lib/brick-shelltest.h index ccb4fb4..cc7135f 100644 --- a/test/lib/brick-shelltest.h +++ b/test/lib/brick-shelltest.h @@ -82,7 +82,7 @@ #include "configure.h" /* Timeout for the whole test suite in hours */ -static const unsigned TEST_SUITE_TIMEOUT = 4; +static const time_t TEST_SUITE_TIMEOUT = 4; #ifndef BRICK_SHELLTEST_H #define BRICK_SHELLTEST_H

1 0

master - tests: use more SKIP
by Zdenek Kabelac 27 Oct '15

27 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=4159680a0ee7eceb… Commit: 4159680a0ee7eceb4d5124c1e4e78b418b12f357 Parent: 76cff10a734a7c1e26b3835ff967dac0b7e46bcb Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 27 15:10:06 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 27 16:00:09 2015 +0100 tests: use more SKIP Speed-up check_lvmpolld. --- test/api/lvtest.sh | 2 ++ test/api/pe_start.sh | 2 ++ test/api/percent.sh | 2 ++ test/api/pytest.sh | 10 +++++----- test/api/thin_percent.sh | 2 ++ test/api/vglist.sh | 2 ++ test/shell/000-basic.sh | 4 ++-- test/shell/covercmd.sh | 4 ++-- test/shell/discards-thin.sh | 4 ++-- test/shell/dmeventd-restart.sh | 4 ++-- test/shell/dumpconfig.sh | 4 ++-- test/shell/error-usage.sh | 4 ++-- test/shell/fsadm.sh | 3 +-- test/shell/inconsistent-metadata.sh | 4 ++-- test/shell/listings.sh | 4 ++-- test/shell/lock-blocking.sh | 5 ++--- test/shell/lock-parallel.sh | 4 ++-- test/shell/lvchange-cache.sh | 4 ++-- test/shell/lvchange-mirror.sh | 9 ++++----- test/shell/lvchange-partial-raid10.sh | 4 ++-- test/shell/lvchange-partial.sh | 4 ++-- test/shell/lvchange-raid.sh | 2 ++ test/shell/lvchange-raid10.sh | 2 -- test/shell/lvchange-raid456.sh | 2 -- test/shell/lvchange-syncaction-raid.sh | 4 ++-- test/shell/lvchange-thin.sh | 4 ++-- test/shell/lvconvert-cache-raid.sh | 4 ++-- test/shell/lvconvert-cache-smq.sh | 4 ++-- test/shell/lvconvert-cache-thin.sh | 4 ++-- test/shell/lvconvert-cache.sh | 4 ++-- test/shell/lvconvert-raid-allocation.sh | 4 ++-- test/shell/lvconvert-raid.sh | 5 +++-- test/shell/lvconvert-raid10.sh | 4 ++-- test/shell/lvconvert-raid456.sh | 4 ++-- test/shell/lvconvert-repair-dmeventd.sh | 4 ++-- test/shell/lvconvert-repair-raid-dmeventd.sh | 4 ++-- test/shell/lvconvert-repair-raid.sh | 4 ++-- test/shell/lvconvert-repair-snapshot.sh | 4 ++-- test/shell/lvconvert-repair-thin-raid.sh | 3 ++- test/shell/lvconvert-repair-thin.sh | 4 ++-- test/shell/lvconvert-repair-transient-dmeventd.sh | 4 ++-- test/shell/lvconvert-repair-transient.sh | 4 ++-- test/shell/lvconvert-snapshot.sh | 4 ++-- test/shell/lvconvert-thin-external.sh | 4 ++-- test/shell/lvconvert-thin-raid.sh | 4 ++-- test/shell/lvconvert-thin.sh | 4 ++-- test/shell/lvcreate-cache-raid.sh | 4 ++-- test/shell/lvcreate-cache.sh | 4 ++-- test/shell/lvcreate-large-raid.sh | 4 ++-- test/shell/lvcreate-large-raid10.sh | 5 ++--- test/shell/lvcreate-large.sh | 4 ++-- test/shell/lvcreate-mirror.sh | 4 ++-- test/shell/lvcreate-missing.sh | 4 ++-- test/shell/lvcreate-operation.sh | 4 ++-- test/shell/lvcreate-pvtags.sh | 4 ++-- test/shell/lvcreate-raid.sh | 4 ++-- test/shell/lvcreate-raid10.sh | 4 ++-- test/shell/lvcreate-repair.sh | 4 ++-- test/shell/lvcreate-signature-wiping.sh | 4 ++-- test/shell/lvcreate-small-snap.sh | 4 ++-- test/shell/lvcreate-striped-mirror.sh | 4 ++-- test/shell/lvcreate-thin-big.sh | 4 ++-- test/shell/lvcreate-thin-external-size.sh | 4 ++-- test/shell/lvcreate-thin-external.sh | 4 ++-- test/shell/lvcreate-thin-power2.sh | 4 ++-- test/shell/lvcreate-thin-snap.sh | 4 ++-- test/shell/lvcreate-thin.sh | 4 ++-- test/shell/lvextend-percent-extents.sh | 4 ++-- test/shell/lvextend-snapshot-dmeventd.sh | 4 ++-- test/shell/lvextend-snapshot-policy.sh | 4 ++-- test/shell/lvextend-thin-full.sh | 4 ++-- test/shell/lvextend-thin-metadata-dmeventd.sh | 4 ++-- test/shell/lvextend-thin.sh | 4 ++-- test/shell/lvm-init.sh | 4 ++-- test/shell/lvmcache-exercise.sh | 4 ++-- test/shell/lvmetad-no-cluster.sh | 3 --- test/shell/lvresize-mirror.sh | 4 ++-- test/shell/lvresize-raid.sh | 4 ++-- test/shell/lvresize-raid10.sh | 4 ++-- test/shell/lvresize-rounding.sh | 4 ++-- test/shell/lvresize-thin-external-origin.sh | 4 ++-- test/shell/lvresize-thin-metadata.sh | 4 ++-- test/shell/lvresize-usage.sh | 4 ++-- test/shell/lvs-cache.sh | 4 ++-- test/shell/mda-rollback.sh | 4 ++-- test/shell/mdata-strings.sh | 4 ++-- test/shell/metadata-balance.sh | 4 ++-- test/shell/metadata-dirs.sh | 4 ++-- test/shell/metadata.sh | 4 ++-- test/shell/mirror-vgreduce-removemissing.sh | 3 +-- test/shell/name-mangling.sh | 10 +++++----- test/shell/nomda-missing.sh | 4 ++-- test/shell/nomda-restoremissing.sh | 4 ++-- test/shell/orphan-ondisk.sh | 4 ++-- test/shell/pool-labels.sh | 8 ++++---- test/shell/process-each-duplicate-vgnames.sh | 3 +-- test/shell/process-each-lv.sh | 1 + test/shell/process-each-pv-nomda-all.sh | 1 + test/shell/process-each-pv-nomda.sh | 1 + test/shell/process-each-pv.sh | 1 + test/shell/process-each-pvresize.sh | 3 +-- test/shell/profiles-thin.sh | 4 ++-- test/shell/profiles.sh | 4 ++-- test/shell/pv-duplicate-uuid.sh | 4 ++-- test/shell/pv-duplicate.sh | 4 ++-- test/shell/pv-min-size.sh | 4 ++-- test/shell/pvchange-usage.sh | 4 ++-- test/shell/pvcreate-bootloaderarea.sh | 3 +-- test/shell/pvcreate-ff.sh | 4 ++-- test/shell/pvcreate-metadata0.sh | 4 ++-- test/shell/pvcreate-operation-md.sh | 4 ++-- test/shell/pvcreate-operation.sh | 4 ++-- test/shell/pvcreate-usage.sh | 3 +-- test/shell/pvmove-cache-segtypes.sh | 2 +- test/shell/pvmove-raid-segtypes.sh | 2 +- test/shell/pvmove-resume-1.sh | 4 ++-- test/shell/pvmove-resume-2.sh | 4 ++-- test/shell/pvmove-resume-multiseg.sh | 4 ++-- test/shell/pvmove-thin-segtypes.sh | 3 ++- test/shell/pvremove-thin.sh | 4 ++-- test/shell/pvremove-usage.sh | 4 ++-- test/shell/pvremove-warnings.sh | 4 ++-- test/shell/read-ahead.sh | 3 +-- test/shell/select-report.sh | 4 ++-- test/shell/select-tools-thin.sh | 4 ++-- test/shell/select-tools.sh | 4 ++-- test/shell/snapshot-autoumount-dmeventd.sh | 4 ++-- test/shell/snapshot-cluster.sh | 4 ++-- test/shell/snapshot-lvm1.sh | 4 ++-- test/shell/snapshot-maxsize.sh | 4 ++-- test/shell/snapshot-reactivate.sh | 5 +++-- test/shell/snapshot-remove-dmsetup.sh | 8 ++++---- test/shell/snapshot-rename.sh | 4 ++-- test/shell/snapshot-usage-exa.sh | 4 ++-- test/shell/snapshot-usage.sh | 4 ++-- test/shell/stray-device-node.sh | 4 ++-- test/shell/system_id.sh | 3 +-- test/shell/tags.sh | 4 ++-- test/shell/test-partition.sh | 4 ++-- test/shell/thin-autoumount-dmeventd.sh | 4 ++-- test/shell/thin-defaults.sh | 4 ++-- test/shell/thin-restore.sh | 4 ++-- test/shell/thin-volume-list.sh | 4 ++-- test/shell/topology-support.sh | 4 ++-- test/shell/unknown-segment.sh | 4 ++-- test/shell/unlost-pv.sh | 4 ++-- test/shell/vg-name-from-env.sh | 3 +-- test/shell/vgcfgbackup-usage.sh | 4 ++-- test/shell/vgchange-many.sh | 4 ++-- test/shell/vgchange-maxlv.sh | 4 ++-- test/shell/vgchange-partial.sh | 4 ++-- test/shell/vgchange-sysinit.sh | 6 +++--- test/shell/vgchange-usage.sh | 3 +-- test/shell/vgck.sh | 4 ++-- test/shell/vgcreate-many-pvs.sh | 4 ++-- test/shell/vgcreate-usage.sh | 3 +-- test/shell/vgextend-restoremissing.sh | 4 ++-- test/shell/vgextend-usage.sh | 4 ++-- test/shell/vgimportclone.sh | 4 ++-- test/shell/vgmerge-operation.sh | 3 +-- test/shell/vgmerge-usage.sh | 3 +-- test/shell/vgreduce-removemissing-snapshot.sh | 7 ++++--- test/shell/vgreduce-usage.sh | 4 ++-- test/shell/vgremove-corrupt-vg.sh | 4 ++-- test/shell/vgrename-usage.sh | 4 ++-- test/shell/vgsplit-operation.sh | 4 ++-- test/shell/vgsplit-raid.sh | 3 +-- test/shell/vgsplit-stacked.sh | 4 ++-- test/shell/vgsplit-thin.sh | 4 ++-- test/shell/vgsplit-usage.sh | 4 ++-- test/shell/zero-usage.sh | 4 ++-- 171 files changed, 333 insertions(+), 336 deletions(-) diff --git a/test/api/lvtest.sh b/test/api/lvtest.sh index 5fe6f57..aacceba 100644 --- a/test/api/lvtest.sh +++ b/test/api/lvtest.sh @@ -11,6 +11,8 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + . lib/inittest aux prepare_vg 1 diff --git a/test/api/pe_start.sh b/test/api/pe_start.sh index 2fc118f..6a47ced 100644 --- a/test/api/pe_start.sh +++ b/test/api/pe_start.sh @@ -11,6 +11,8 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + . lib/inittest aux prepare_devs 2 diff --git a/test/api/percent.sh b/test/api/percent.sh index a9b7b90..d574080 100644 --- a/test/api/percent.sh +++ b/test/api/percent.sh @@ -11,6 +11,8 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + . lib/inittest aux kernel_at_least 2 6 33 || skip diff --git a/test/api/pytest.sh b/test/api/pytest.sh index 3b7c5b6..7ede416 100644 --- a/test/api/pytest.sh +++ b/test/api/pytest.sh @@ -11,8 +11,13 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMETAD=1 +SKIP_WITH_CLVMD=1 + . lib/inittest +aux prepare_dmeventd + # # TODO: # lvm2app is not yet capable to respect many lvm.conf options @@ -29,11 +34,6 @@ python_lib=$(find $abs_top_builddir -name lvm.so) # Unable to test python bindings if library not available test -n "$python_lib" || skip -test -e LOCAL_CLVMD && skip -test -e LOCAL_LVMETAD && skip - -aux prepare_dmeventd - #If you change this change the unit test case too. aux prepare_pvs 6 diff --git a/test/api/thin_percent.sh b/test/api/thin_percent.sh index fd64e3b..cd95e4f 100644 --- a/test/api/thin_percent.sh +++ b/test/api/thin_percent.sh @@ -11,6 +11,8 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest diff --git a/test/api/vglist.sh b/test/api/vglist.sh index 421705b..66cf7bf 100644 --- a/test/api/vglist.sh +++ b/test/api/vglist.sh @@ -9,6 +9,8 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + . lib/inittest aux prepare_vg 2 diff --git a/test/shell/000-basic.sh b/test/shell/000-basic.sh index 582c5ef..89b2178 100644 --- a/test/shell/000-basic.sh +++ b/test/shell/000-basic.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest lvm version diff --git a/test/shell/covercmd.sh b/test/shell/covercmd.sh index 5e47bd9..db84524 100644 --- a/test/shell/covercmd.sh +++ b/test/shell/covercmd.sh @@ -14,9 +14,9 @@ # to improve code coverage # -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 5 get_devs diff --git a/test/shell/discards-thin.sh b/test/shell/discards-thin.sh index fbd0d87..95a5880 100644 --- a/test/shell/discards-thin.sh +++ b/test/shell/discards-thin.sh @@ -13,12 +13,12 @@ # test support of thin discards # +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - # # Main # diff --git a/test/shell/dmeventd-restart.sh b/test/shell/dmeventd-restart.sh index a64509e..c9f4dc0 100644 --- a/test/shell/dmeventd-restart.sh +++ b/test/shell/dmeventd-restart.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_dmeventd diff --git a/test/shell/dumpconfig.sh b/test/shell/dumpconfig.sh index 0d33761..312854e 100644 --- a/test/shell/dumpconfig.sh +++ b/test/shell/dumpconfig.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest flatten() { cat > flatten.config diff --git a/test/shell/error-usage.sh b/test/shell/error-usage.sh index f617b61..ba9afe8 100644 --- a/test/shell/error-usage.sh +++ b/test/shell/error-usage.sh @@ -11,9 +11,9 @@ # Basic usage of zero target -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest which md5sum || skip diff --git a/test/shell/fsadm.sh b/test/shell/fsadm.sh index 12f1593..f539981 100644 --- a/test/shell/fsadm.sh +++ b/test/shell/fsadm.sh @@ -10,11 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Exercise fsadm filesystem resize' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_vg 1 100 # set to "skip" to avoid testing given fs and test warning result diff --git a/test/shell/inconsistent-metadata.sh b/test/shell/inconsistent-metadata.sh index a606330..6fc09ee 100644 --- a/test/shell/inconsistent-metadata.sh +++ b/test/shell/inconsistent-metadata.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 3 12 diff --git a/test/shell/listings.sh b/test/shell/listings.sh index d018510..3f1f302 100644 --- a/test/shell/listings.sh +++ b/test/shell/listings.sh @@ -13,9 +13,9 @@ # tests functionality of lvs, pvs, vgs, *display tools # -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 5 get_devs diff --git a/test/shell/lock-blocking.sh b/test/shell/lock-blocking.sh index 5c004c9..35bf793 100644 --- a/test/shell/lock-blocking.sh +++ b/test/shell/lock-blocking.sh @@ -10,12 +10,11 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='test some blocking / non-blocking multi-vg operations' +SKIP_WITH_CLVMD=1 +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip -test -e LOCAL_CLVMD && skip - aux prepare_devs 3 pvcreate "$dev1" "$dev2" vgcreate $vg "$dev1" "$dev2" diff --git a/test/shell/lock-parallel.sh b/test/shell/lock-parallel.sh index ebe6336..b931e4f 100644 --- a/test/shell/lock-parallel.sh +++ b/test/shell/lock-parallel.sh @@ -12,9 +12,9 @@ # Test parallel use of lvm commands and check locks aren't dropped # RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=1049296 -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest which mkfs.ext3 || skip which fsck || skip diff --git a/test/shell/lvchange-cache.sh b/test/shell/lvchange-cache.sh index 3efd762..4a5c8af 100644 --- a/test/shell/lvchange-cache.sh +++ b/test/shell/lvchange-cache.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_cache 1 3 0 || skip aux prepare_vg 3 diff --git a/test/shell/lvchange-mirror.sh b/test/shell/lvchange-mirror.sh index bd4c868..a828251 100644 --- a/test/shell/lvchange-mirror.sh +++ b/test/shell/lvchange-mirror.sh @@ -9,13 +9,12 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest - -test -e LOCAL_LVMPOLLD && skip - # FIXME RESYNC doesn't work in cluster with exclusive activation # seriously broken! -test -e LOCAL_CLVMD && skip +SKIP_WITH_CLVMD=1 +SKIP_WITH_LVMPOLLD=1 + +. lib/inittest aux prepare_dmeventd aux prepare_vg 3 diff --git a/test/shell/lvchange-partial-raid10.sh b/test/shell/lvchange-partial-raid10.sh index d83d193..0a20a01 100644 --- a/test/shell/lvchange-partial-raid10.sh +++ b/test/shell/lvchange-partial-raid10.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_raid 1 3 2 || skip diff --git a/test/shell/lvchange-partial.sh b/test/shell/lvchange-partial.sh index 7b1f68d..6ce46eb 100644 --- a/test/shell/lvchange-partial.sh +++ b/test/shell/lvchange-partial.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 4 diff --git a/test/shell/lvchange-raid.sh b/test/shell/lvchange-raid.sh index 4ce776f..c572836 100644 --- a/test/shell/lvchange-raid.sh +++ b/test/shell/lvchange-raid.sh @@ -9,6 +9,8 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + . lib/inittest # Writemostly has been in every version since the begining diff --git a/test/shell/lvchange-raid10.sh b/test/shell/lvchange-raid10.sh index 574ac7d..0d17928 100644 --- a/test/shell/lvchange-raid10.sh +++ b/test/shell/lvchange-raid10.sh @@ -13,8 +13,6 @@ TEST_RAID=raid10 . shell/lvchange-raid.sh -test -e LOCAL_LVMPOLLD && skip - aux have_raid 1 5 2 || skip run_types raid10 -m 1 -i 2 "$dev1" "$dev2" "$dev3" "$dev4" diff --git a/test/shell/lvchange-raid456.sh b/test/shell/lvchange-raid456.sh index bdce2e6..b1dd1db 100644 --- a/test/shell/lvchange-raid456.sh +++ b/test/shell/lvchange-raid456.sh @@ -13,8 +13,6 @@ TEST_RAID=raid456 . shell/lvchange-raid.sh -test -e LOCAL_LVMPOLLD && skip - aux raid456_replace_works || skip aux have_raid 1 5 2 || skip diff --git a/test/shell/lvchange-syncaction-raid.sh b/test/shell/lvchange-syncaction-raid.sh index 558157b..704c5d8 100644 --- a/test/shell/lvchange-syncaction-raid.sh +++ b/test/shell/lvchange-syncaction-raid.sh @@ -11,9 +11,9 @@ # test activation race for raid's --syncaction check -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # Proper mismatch count 1.5.2+ upstream, 1.3.5 < x < 1.4.0 in RHEL6 aux have_raid 1 3 5 && diff --git a/test/shell/lvchange-thin.sh b/test/shell/lvchange-thin.sh index b66c5ee..cad4873 100644 --- a/test/shell/lvchange-thin.sh +++ b/test/shell/lvchange-thin.sh @@ -9,12 +9,12 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux have_thin 1 0 0 || skip aux prepare_pvs 3 diff --git a/test/shell/lvconvert-cache-raid.sh b/test/shell/lvconvert-cache-raid.sh index 284ffbe..b0d8da2 100644 --- a/test/shell/lvconvert-cache-raid.sh +++ b/test/shell/lvconvert-cache-raid.sh @@ -11,9 +11,9 @@ # Exercise usage of stacked cache volume using raid volume -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_cache 1 3 0 || skip aux have_raid 1 0 0 || skip diff --git a/test/shell/lvconvert-cache-smq.sh b/test/shell/lvconvert-cache-smq.sh index cdc7838..75f447f 100644 --- a/test/shell/lvconvert-cache-smq.sh +++ b/test/shell/lvconvert-cache-smq.sh @@ -11,9 +11,9 @@ # Exercise conversion of cache and cache pool -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_cache 1 8 0 || skip diff --git a/test/shell/lvconvert-cache-thin.sh b/test/shell/lvconvert-cache-thin.sh index 12b8dfe..ace4eb5 100644 --- a/test/shell/lvconvert-cache-thin.sh +++ b/test/shell/lvconvert-cache-thin.sh @@ -11,9 +11,9 @@ # Exercise usage of stacked cache volume used in thin pool volumes -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_cache 1 3 0 || skip aux have_thin 1 0 0 || skip diff --git a/test/shell/lvconvert-cache.sh b/test/shell/lvconvert-cache.sh index 3c94ee3..5a38d9c 100644 --- a/test/shell/lvconvert-cache.sh +++ b/test/shell/lvconvert-cache.sh @@ -11,9 +11,9 @@ # Exercise conversion of cache and cache pool -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_cache 1 3 0 || skip diff --git a/test/shell/lvconvert-raid-allocation.sh b/test/shell/lvconvert-raid-allocation.sh index f118d9c..01a71cb 100644 --- a/test/shell/lvconvert-raid-allocation.sh +++ b/test/shell/lvconvert-raid-allocation.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_raid 1 3 0 || skip diff --git a/test/shell/lvconvert-raid.sh b/test/shell/lvconvert-raid.sh index 6529ee2..0b56fc3 100644 --- a/test/shell/lvconvert-raid.sh +++ b/test/shell/lvconvert-raid.sh @@ -10,12 +10,13 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA # disable lvmetad logging as it bogs down test systems + +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_LVMETAD_DEBUG_OPTS=${LVM_TEST_LVMETAD_DEBUG_OPTS-} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - get_image_pvs() { local d local images diff --git a/test/shell/lvconvert-raid10.sh b/test/shell/lvconvert-raid10.sh index 244a79e..aa8ceb1 100644 --- a/test/shell/lvconvert-raid10.sh +++ b/test/shell/lvconvert-raid10.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest get_image_pvs() { local d diff --git a/test/shell/lvconvert-raid456.sh b/test/shell/lvconvert-raid456.sh index 0fb5a5b..8f2518c 100644 --- a/test/shell/lvconvert-raid456.sh +++ b/test/shell/lvconvert-raid456.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest get_image_pvs() { local d diff --git a/test/shell/lvconvert-repair-dmeventd.sh b/test/shell/lvconvert-repair-dmeventd.sh index 2acf461..308ea4b 100644 --- a/test/shell/lvconvert-repair-dmeventd.sh +++ b/test/shell/lvconvert-repair-dmeventd.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest which mkfs.ext2 || skip aux mirror_recovery_works || skip diff --git a/test/shell/lvconvert-repair-raid-dmeventd.sh b/test/shell/lvconvert-repair-raid-dmeventd.sh index a5a7a53..be61618 100644 --- a/test/shell/lvconvert-repair-raid-dmeventd.sh +++ b/test/shell/lvconvert-repair-raid-dmeventd.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest which mkfs.ext3 || skip aux have_raid 1 3 0 || skip diff --git a/test/shell/lvconvert-repair-raid.sh b/test/shell/lvconvert-repair-raid.sh index 580516e..f1460ab 100644 --- a/test/shell/lvconvert-repair-raid.sh +++ b/test/shell/lvconvert-repair-raid.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_raid 1 3 0 || skip aux raid456_replace_works || skip diff --git a/test/shell/lvconvert-repair-snapshot.sh b/test/shell/lvconvert-repair-snapshot.sh index 73f2f88..67b58a9 100644 --- a/test/shell/lvconvert-repair-snapshot.sh +++ b/test/shell/lvconvert-repair-snapshot.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 5 aux lvmconf 'allocation/maximise_cling = 0' \ diff --git a/test/shell/lvconvert-repair-thin-raid.sh b/test/shell/lvconvert-repair-thin-raid.sh index 04bc5c7..0eb80fa 100644 --- a/test/shell/lvconvert-repair-thin-raid.sh +++ b/test/shell/lvconvert-repair-thin-raid.sh @@ -12,9 +12,10 @@ # Test repairing of broken thin pool on raid +SKIP_WITH_LVMPOLLD=1 + . lib/inittest -test -e LOCAL_LVMPOLLD && skip aux have_thin 1 0 0 || skip aux have_raid 1 4 0 || skip diff --git a/test/shell/lvconvert-repair-thin.sh b/test/shell/lvconvert-repair-thin.sh index c699e12..90f1480 100644 --- a/test/shell/lvconvert-repair-thin.sh +++ b/test/shell/lvconvert-repair-thin.sh @@ -12,9 +12,9 @@ # Test repairing of broken thin pool metadata -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest which mkfs.ext2 || skip diff --git a/test/shell/lvconvert-repair-transient-dmeventd.sh b/test/shell/lvconvert-repair-transient-dmeventd.sh index 30fd47c..17778fb 100644 --- a/test/shell/lvconvert-repair-transient-dmeventd.sh +++ b/test/shell/lvconvert-repair-transient-dmeventd.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_dmeventd aux mirror_recovery_works || skip diff --git a/test/shell/lvconvert-repair-transient.sh b/test/shell/lvconvert-repair-transient.sh index 52e1825..ec17f92 100644 --- a/test/shell/lvconvert-repair-transient.sh +++ b/test/shell/lvconvert-repair-transient.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux mirror_recovery_works || skip aux prepare_vg 5 diff --git a/test/shell/lvconvert-snapshot.sh b/test/shell/lvconvert-snapshot.sh index ad500e2..b4cb443 100644 --- a/test/shell/lvconvert-snapshot.sh +++ b/test/shell/lvconvert-snapshot.sh @@ -12,9 +12,9 @@ # Test various supported conversion of snapshot -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 1 diff --git a/test/shell/lvconvert-thin-external.sh b/test/shell/lvconvert-thin-external.sh index 479d23c..4382f1a 100644 --- a/test/shell/lvconvert-thin-external.sh +++ b/test/shell/lvconvert-thin-external.sh @@ -12,12 +12,12 @@ # Test conversion to thin external origin +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - which mkfs.ext2 || skip which fsck || skip diff --git a/test/shell/lvconvert-thin-raid.sh b/test/shell/lvconvert-thin-raid.sh index bd554b2..e9a5917 100644 --- a/test/shell/lvconvert-thin-raid.sh +++ b/test/shell/lvconvert-thin-raid.sh @@ -10,12 +10,12 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux have_thin 1 0 0 || skip aux have_raid 1 4 0 || skip diff --git a/test/shell/lvconvert-thin.sh b/test/shell/lvconvert-thin.sh index b4b3a72..12d4331 100644 --- a/test/shell/lvconvert-thin.sh +++ b/test/shell/lvconvert-thin.sh @@ -10,12 +10,12 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - prepare_lvs() { lvremove -f $vg lvcreate -L10M -n $lv1 $vg diff --git a/test/shell/lvcreate-cache-raid.sh b/test/shell/lvcreate-cache-raid.sh index 1c3c327..6e8b3cc 100644 --- a/test/shell/lvcreate-cache-raid.sh +++ b/test/shell/lvcreate-cache-raid.sh @@ -11,9 +11,9 @@ # Exercise creation of cache and raids -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_cache 1 3 0 || skip aux have_raid 1 0 0 || skip diff --git a/test/shell/lvcreate-cache.sh b/test/shell/lvcreate-cache.sh index fce88f3..b3e9d00 100644 --- a/test/shell/lvcreate-cache.sh +++ b/test/shell/lvcreate-cache.sh @@ -14,9 +14,9 @@ # Full CLI uses --type # Shorthand CLI uses -H | --cache -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_cache 1 3 0 || skip diff --git a/test/shell/lvcreate-large-raid.sh b/test/shell/lvcreate-large-raid.sh index c59eded..46f76b9 100644 --- a/test/shell/lvcreate-large-raid.sh +++ b/test/shell/lvcreate-large-raid.sh @@ -11,9 +11,9 @@ # 'Exercise some lvcreate diagnostics' -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # FIXME update test to make something useful on <16T aux can_use_16T || skip diff --git a/test/shell/lvcreate-large-raid10.sh b/test/shell/lvcreate-large-raid10.sh index 47dda2b..2bc1581 100644 --- a/test/shell/lvcreate-large-raid10.sh +++ b/test/shell/lvcreate-large-raid10.sh @@ -11,13 +11,12 @@ # 'Exercise some lvcreate diagnostics' -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # FIXME update test to make something useful on <16T aux can_use_16T || skip - aux have_raid 1 3 0 || skip aux prepare_vg 5 diff --git a/test/shell/lvcreate-large.sh b/test/shell/lvcreate-large.sh index e547fcc..f51c9b6 100644 --- a/test/shell/lvcreate-large.sh +++ b/test/shell/lvcreate-large.sh @@ -11,9 +11,9 @@ # 'Exercise some lvcreate diagnostics' -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # FIXME update test to make something useful on <16T aux can_use_16T || skip diff --git a/test/shell/lvcreate-mirror.sh b/test/shell/lvcreate-mirror.sh index 275c036..5739999 100644 --- a/test/shell/lvcreate-mirror.sh +++ b/test/shell/lvcreate-mirror.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 5 80 aux lvmconf 'allocation/maximise_cling = 0' \ diff --git a/test/shell/lvcreate-missing.sh b/test/shell/lvcreate-missing.sh index 012962f..baef6ff 100644 --- a/test/shell/lvcreate-missing.sh +++ b/test/shell/lvcreate-missing.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 2 diff --git a/test/shell/lvcreate-operation.sh b/test/shell/lvcreate-operation.sh index 3947bf8..bec8502 100644 --- a/test/shell/lvcreate-operation.sh +++ b/test/shell/lvcreate-operation.sh @@ -11,9 +11,9 @@ # 'Exercise some lvcreate diagnostics' -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest cleanup_lvs() { lvremove -ff $vg diff --git a/test/shell/lvcreate-pvtags.sh b/test/shell/lvcreate-pvtags.sh index 99a7f07..2a60f69 100644 --- a/test/shell/lvcreate-pvtags.sh +++ b/test/shell/lvcreate-pvtags.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 3 aux lvmconf 'allocation/maximise_cling = 0' \ diff --git a/test/shell/lvcreate-raid.sh b/test/shell/lvcreate-raid.sh index f5e25c6..35cd867 100644 --- a/test/shell/lvcreate-raid.sh +++ b/test/shell/lvcreate-raid.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest lv_devices() { test $3 -eq $(get lv_devices $1/$2 | wc -w) diff --git a/test/shell/lvcreate-raid10.sh b/test/shell/lvcreate-raid10.sh index 884653b..0c8ae38 100644 --- a/test/shell/lvcreate-raid10.sh +++ b/test/shell/lvcreate-raid10.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest lv_devices() { test $3 -eq $(get lv_devices $1/$2 | wc -w) diff --git a/test/shell/lvcreate-repair.sh b/test/shell/lvcreate-repair.sh index 469bf39..4bcf782 100644 --- a/test/shell/lvcreate-repair.sh +++ b/test/shell/lvcreate-repair.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 3 diff --git a/test/shell/lvcreate-signature-wiping.sh b/test/shell/lvcreate-signature-wiping.sh index d2d23c5..e4bfdb0 100644 --- a/test/shell/lvcreate-signature-wiping.sh +++ b/test/shell/lvcreate-signature-wiping.sh @@ -11,9 +11,9 @@ # 'Exercise signature wiping during lvcreate' -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest init_lv_() { mkswap "$DM_DEV_DIR/$vg/$lv1" diff --git a/test/shell/lvcreate-small-snap.sh b/test/shell/lvcreate-small-snap.sh index b5610c7..e3f5189 100644 --- a/test/shell/lvcreate-small-snap.sh +++ b/test/shell/lvcreate-small-snap.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs diff --git a/test/shell/lvcreate-striped-mirror.sh b/test/shell/lvcreate-striped-mirror.sh index 776dbf7..090fb73 100644 --- a/test/shell/lvcreate-striped-mirror.sh +++ b/test/shell/lvcreate-striped-mirror.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 9 diff --git a/test/shell/lvcreate-thin-big.sh b/test/shell/lvcreate-thin-big.sh index dbd0986..797bc5b 100644 --- a/test/shell/lvcreate-thin-big.sh +++ b/test/shell/lvcreate-thin-big.sh @@ -13,12 +13,12 @@ # test currently needs to drop # 'return NULL' in _lv_create_an_lv after log_error("Can't create %s without using " +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux have_thin 1 0 0 || skip # Test --poolmetadatasize range diff --git a/test/shell/lvcreate-thin-external-size.sh b/test/shell/lvcreate-thin-external-size.sh index e2f49c4..38b8a3d 100644 --- a/test/shell/lvcreate-thin-external-size.sh +++ b/test/shell/lvcreate-thin-external-size.sh @@ -12,12 +12,12 @@ # Test unaligned size of external origin and thin pool chunk size +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - which cmp || skip # diff --git a/test/shell/lvcreate-thin-external.sh b/test/shell/lvcreate-thin-external.sh index 8507836..c37a898 100644 --- a/test/shell/lvcreate-thin-external.sh +++ b/test/shell/lvcreate-thin-external.sh @@ -12,12 +12,12 @@ # Test creation of thin snapshots using external origin +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - which mkfs.ext2 || skip which fsck || skip diff --git a/test/shell/lvcreate-thin-power2.sh b/test/shell/lvcreate-thin-power2.sh index 4d3d175..c6bf4a0 100644 --- a/test/shell/lvcreate-thin-power2.sh +++ b/test/shell/lvcreate-thin-power2.sh @@ -13,12 +13,12 @@ # test support for non-power-of-2 thin chunk size # +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - # # Main # diff --git a/test/shell/lvcreate-thin-snap.sh b/test/shell/lvcreate-thin-snap.sh index 76929cc..66a815b 100644 --- a/test/shell/lvcreate-thin-snap.sh +++ b/test/shell/lvcreate-thin-snap.sh @@ -10,12 +10,12 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - check_lv_field_modules_() { mod=$1 diff --git a/test/shell/lvcreate-thin.sh b/test/shell/lvcreate-thin.sh index 56d7e35..236bc27 100644 --- a/test/shell/lvcreate-thin.sh +++ b/test/shell/lvcreate-thin.sh @@ -13,12 +13,12 @@ # test currently needs to drop # 'return NULL' in _lv_create_an_lv after log_error("Can't create %s without using " +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - check_lv_field_modules_() { mod=$1 diff --git a/test/shell/lvextend-percent-extents.sh b/test/shell/lvextend-percent-extents.sh index ec98e59..0460fcb 100644 --- a/test/shell/lvextend-percent-extents.sh +++ b/test/shell/lvextend-percent-extents.sh @@ -11,9 +11,9 @@ # 'Check extents percentage arguments' -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 128 vgcreate $vg $(cat DEVICES) diff --git a/test/shell/lvextend-snapshot-dmeventd.sh b/test/shell/lvextend-snapshot-dmeventd.sh index aeb18f1..7506fe5 100644 --- a/test/shell/lvextend-snapshot-dmeventd.sh +++ b/test/shell/lvextend-snapshot-dmeventd.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest extend() { lvextend --use-policies --config "activation { snapshot_autoextend_threshold = $1 }" $vg/snap diff --git a/test/shell/lvextend-snapshot-policy.sh b/test/shell/lvextend-snapshot-policy.sh index c76e36e..6eb7dad 100644 --- a/test/shell/lvextend-snapshot-policy.sh +++ b/test/shell/lvextend-snapshot-policy.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest which mkfs.ext2 || skip diff --git a/test/shell/lvextend-thin-full.sh b/test/shell/lvextend-thin-full.sh index 6b1b9e0..31ab729 100644 --- a/test/shell/lvextend-thin-full.sh +++ b/test/shell/lvextend-thin-full.sh @@ -13,12 +13,12 @@ # play with thin-pool resize in corner cases # +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux have_thin 1 0 0 || skip test -n "$LVM_TEST_THIN_RESTORE_CMD" || LVM_TEST_THIN_RESTORE_CMD=$(which thin_restore) || skip diff --git a/test/shell/lvextend-thin-metadata-dmeventd.sh b/test/shell/lvextend-thin-metadata-dmeventd.sh index 9ecd5ae..030991f 100644 --- a/test/shell/lvextend-thin-metadata-dmeventd.sh +++ b/test/shell/lvextend-thin-metadata-dmeventd.sh @@ -11,12 +11,12 @@ # Test autoextension of thin metadata volume +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - meta_percent_() { get lv_field $vg/pool metadata_percent | cut -d. -f1 } diff --git a/test/shell/lvextend-thin.sh b/test/shell/lvextend-thin.sh index a97e93b..ac78a88 100644 --- a/test/shell/lvextend-thin.sh +++ b/test/shell/lvextend-thin.sh @@ -9,12 +9,12 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux have_thin 1 0 0 || skip aux prepare_vg 3 diff --git a/test/shell/lvm-init.sh b/test/shell/lvm-init.sh index cfdd4cc..2e81d42 100644 --- a/test/shell/lvm-init.sh +++ b/test/shell/lvm-init.sh @@ -13,9 +13,9 @@ # tests lvm initialization, and especially negative tests of error paths # -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 5 diff --git a/test/shell/lvmcache-exercise.sh b/test/shell/lvmcache-exercise.sh index f0d3ee9..2c2be52 100644 --- a/test/shell/lvmcache-exercise.sh +++ b/test/shell/lvmcache-exercise.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 5 diff --git a/test/shell/lvmetad-no-cluster.sh b/test/shell/lvmetad-no-cluster.sh index 0421c59..41d7543 100644 --- a/test/shell/lvmetad-no-cluster.sh +++ b/test/shell/lvmetad-no-cluster.sh @@ -14,9 +14,6 @@ SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_CLVMD || skip -test -e LOCAL_LVMPOLLD && skip - aux prepare_vg 2 aux prepare_lvmetad vgs -vv 2> errs diff --git a/test/shell/lvresize-mirror.sh b/test/shell/lvresize-mirror.sh index 423c9ed..13d5aa4 100644 --- a/test/shell/lvresize-mirror.sh +++ b/test/shell/lvresize-mirror.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 5 diff --git a/test/shell/lvresize-raid.sh b/test/shell/lvresize-raid.sh index 333318a..b13bcb2 100644 --- a/test/shell/lvresize-raid.sh +++ b/test/shell/lvresize-raid.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_raid 1 3 0 || skip diff --git a/test/shell/lvresize-raid10.sh b/test/shell/lvresize-raid10.sh index 4fe49f6..fe76340 100644 --- a/test/shell/lvresize-raid10.sh +++ b/test/shell/lvresize-raid10.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_raid 1 3 0 || skip diff --git a/test/shell/lvresize-rounding.sh b/test/shell/lvresize-rounding.sh index 398740e..8f9974a 100644 --- a/test/shell/lvresize-rounding.sh +++ b/test/shell/lvresize-rounding.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 3 22 diff --git a/test/shell/lvresize-thin-external-origin.sh b/test/shell/lvresize-thin-external-origin.sh index 090cb06..540dfe8 100644 --- a/test/shell/lvresize-thin-external-origin.sh +++ b/test/shell/lvresize-thin-external-origin.sh @@ -11,12 +11,12 @@ # Test resize of thin volume with external origin +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux have_thin 1 2 0 || skip # Pretend we miss the external_origin_extend feature diff --git a/test/shell/lvresize-thin-metadata.sh b/test/shell/lvresize-thin-metadata.sh index 82ac898..c097db1 100644 --- a/test/shell/lvresize-thin-metadata.sh +++ b/test/shell/lvresize-thin-metadata.sh @@ -9,12 +9,12 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux have_thin 1 10 0 || skip aux prepare_pvs 3 1256 diff --git a/test/shell/lvresize-usage.sh b/test/shell/lvresize-usage.sh index 00bae14..a682a66 100644 --- a/test/shell/lvresize-usage.sh +++ b/test/shell/lvresize-usage.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 2 80 diff --git a/test/shell/lvs-cache.sh b/test/shell/lvs-cache.sh index 73706b7..072e767 100644 --- a/test/shell/lvs-cache.sh +++ b/test/shell/lvs-cache.sh @@ -14,9 +14,9 @@ # Full CLI uses --type # Shorthand CLI uses -H | --cache -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux have_cache 1 3 0 || skip aux prepare_vg 5 8000 diff --git a/test/shell/mda-rollback.sh b/test/shell/mda-rollback.sh index 5f348fb..80066c2 100644 --- a/test/shell/mda-rollback.sh +++ b/test/shell/mda-rollback.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 3 diff --git a/test/shell/mdata-strings.sh b/test/shell/mdata-strings.sh index 529e599..fda09ec 100644 --- a/test/shell/mdata-strings.sh +++ b/test/shell/mdata-strings.sh @@ -11,9 +11,9 @@ # 'Test for proper escaping of strings in metadata (bz431474)' -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # For udev impossible to create test "$LVM_TEST_DEVDIR" = "/dev" && skip diff --git a/test/shell/metadata-balance.sh b/test/shell/metadata-balance.sh index 869297e..447b4e1 100644 --- a/test/shell/metadata-balance.sh +++ b/test/shell/metadata-balance.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 6 diff --git a/test/shell/metadata-dirs.sh b/test/shell/metadata-dirs.sh index eed6512..04e483d 100644 --- a/test/shell/metadata-dirs.sh +++ b/test/shell/metadata-dirs.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 3 diff --git a/test/shell/metadata.sh b/test/shell/metadata.sh index 8e3503f..ebf341c 100644 --- a/test/shell/metadata.sh +++ b/test/shell/metadata.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 5 diff --git a/test/shell/mirror-vgreduce-removemissing.sh b/test/shell/mirror-vgreduce-removemissing.sh index 40a6503..36969ea 100644 --- a/test/shell/mirror-vgreduce-removemissing.sh +++ b/test/shell/mirror-vgreduce-removemissing.sh @@ -11,14 +11,13 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description="ensure that 'vgreduce --removemissing' works on mirrored LV" +SKIP_WITH_LVMPOLLD=1 # disable lvmetad logging as it bogs down test systems export LVM_TEST_LVMETAD_DEBUG_OPTS=${LVM_TEST_LVMETAD_DEBUG_OPTS-} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - lv_is_on_ () { local lv=$vg/$1 diff --git a/test/shell/name-mangling.sh b/test/shell/name-mangling.sh index 6bdf4a1..c9609fe 100644 --- a/test/shell/name-mangling.sh +++ b/test/shell/name-mangling.sh @@ -9,13 +9,13 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest - # This test is not using any lvm command # so skip duplicate CLMVD and lvmetad test -test -e LOCAL_CLVMD && skip -test -e LOCAL_LVMETAD && skip -test -e LOCAL_LVMPOLLD && skip +SKIP_WITH_CLVMD=1 +SKIP_WITH_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 + +. lib/inittest CHARACTER_WHITELIST="ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789#+-.:=@_" FAIL_MIXED_STR="contains mixed mangled and unmangled characters" diff --git a/test/shell/nomda-missing.sh b/test/shell/nomda-missing.sh index 39c2ac9..77e81eb 100644 --- a/test/shell/nomda-missing.sh +++ b/test/shell/nomda-missing.sh @@ -10,9 +10,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 4 pvcreate "$dev1" "$dev2" diff --git a/test/shell/nomda-restoremissing.sh b/test/shell/nomda-restoremissing.sh index 9380c8d..39246dd 100644 --- a/test/shell/nomda-restoremissing.sh +++ b/test/shell/nomda-restoremissing.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 3 diff --git a/test/shell/orphan-ondisk.sh b/test/shell/orphan-ondisk.sh index f4e18f4..6b2254e 100644 --- a/test/shell/orphan-ondisk.sh +++ b/test/shell/orphan-ondisk.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 2 vgreduce $vg "$dev1" 2>&1 | not grep -i 'parse error' diff --git a/test/shell/pool-labels.sh b/test/shell/pool-labels.sh index 70d5a64..395cd20 100644 --- a/test/shell/pool-labels.sh +++ b/test/shell/pool-labels.sh @@ -9,11 +9,11 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest - # lvmetad does not handle pool labels so skip test. -test ! -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +SKIP_WITH_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 + +. lib/inittest env printf "" || skip # skip if printf is not available diff --git a/test/shell/process-each-duplicate-vgnames.sh b/test/shell/process-each-duplicate-vgnames.sh index f2b4765..90d3e23 100644 --- a/test/shell/process-each-duplicate-vgnames.sh +++ b/test/shell/process-each-duplicate-vgnames.sh @@ -6,11 +6,10 @@ # of the GNU General Public License v.2. test_description='Test vgs with duplicate vg names' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_devs 2 pvcreate "$dev1" diff --git a/test/shell/process-each-lv.sh b/test/shell/process-each-lv.sh index 52c97d4..8b69271 100644 --- a/test/shell/process-each-lv.sh +++ b/test/shell/process-each-lv.sh @@ -10,6 +10,7 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Exercise toollib process_each_lv' +SKIP_WITH_LVMPOLLD=1 # disable lvmetad logging as it bogs down test systems export LVM_TEST_LVMETAD_DEBUG_OPTS=${LVM_TEST_LVMETAD_DEBUG_OPTS-} diff --git a/test/shell/process-each-pv-nomda-all.sh b/test/shell/process-each-pv-nomda-all.sh index 972bb85..5919a76 100644 --- a/test/shell/process-each-pv-nomda-all.sh +++ b/test/shell/process-each-pv-nomda-all.sh @@ -10,6 +10,7 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Test process_each_pv with zero mda' +SKIP_WITH_LVMPOLLD=1 . lib/inittest diff --git a/test/shell/process-each-pv-nomda.sh b/test/shell/process-each-pv-nomda.sh index 13fddb3..6a42fee 100644 --- a/test/shell/process-each-pv-nomda.sh +++ b/test/shell/process-each-pv-nomda.sh @@ -10,6 +10,7 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Test process_each_pv with zero mda' +SKIP_WITH_LVMPOLLD=1 . lib/inittest diff --git a/test/shell/process-each-pv.sh b/test/shell/process-each-pv.sh index 426c3f1..886b600 100644 --- a/test/shell/process-each-pv.sh +++ b/test/shell/process-each-pv.sh @@ -10,6 +10,7 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Exercise toollib process_each_pv' +SKIP_WITH_LVMPOLLD=1 . lib/inittest diff --git a/test/shell/process-each-pvresize.sh b/test/shell/process-each-pvresize.sh index 493c24b..e3a03d4 100644 --- a/test/shell/process-each-pvresize.sh +++ b/test/shell/process-each-pvresize.sh @@ -10,11 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Exercise toollib process_each_pv' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_devs 14 # diff --git a/test/shell/profiles-thin.sh b/test/shell/profiles-thin.sh index 5940cc6..8f7f1db 100644 --- a/test/shell/profiles-thin.sh +++ b/test/shell/profiles-thin.sh @@ -13,12 +13,12 @@ # test thin profile functionality # +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - DEV_SIZE=32 # check we have thinp support compiled in diff --git a/test/shell/profiles.sh b/test/shell/profiles.sh index 8187129..aaa1dcf 100644 --- a/test/shell/profiles.sh +++ b/test/shell/profiles.sh @@ -13,9 +13,9 @@ # test basic profile functionality # -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest MSG_FAILED_TO_APPLY_CMD_PROFILE="Failed to apply command profile" MSG_IGNORING_INVALID_CMD_PROFILE="Ignoring invalid command profile" diff --git a/test/shell/pv-duplicate-uuid.sh b/test/shell/pv-duplicate-uuid.sh index 7b8b42e..6dbcc75 100644 --- a/test/shell/pv-duplicate-uuid.sh +++ b/test/shell/pv-duplicate-uuid.sh @@ -10,9 +10,9 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA # Test 'Found duplicate' is shown -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 3 diff --git a/test/shell/pv-duplicate.sh b/test/shell/pv-duplicate.sh index ecd6125..99668ee 100644 --- a/test/shell/pv-duplicate.sh +++ b/test/shell/pv-duplicate.sh @@ -11,9 +11,9 @@ # 'Exercise duplicate metadata diagnostics' -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 3 diff --git a/test/shell/pv-min-size.sh b/test/shell/pv-min-size.sh index eb62c73..d0b2c21 100644 --- a/test/shell/pv-min-size.sh +++ b/test/shell/pv-min-size.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # use small default size - 512KB aux lvmconf 'devices/pv_min_size = 512' diff --git a/test/shell/pvchange-usage.sh b/test/shell/pvchange-usage.sh index 8cffb06..ca78168 100644 --- a/test/shell/pvchange-usage.sh +++ b/test/shell/pvchange-usage.sh @@ -11,9 +11,9 @@ # 'Test pvchange option values' -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest check_changed_uuid_() { test "$1" != "$(get pv_field "$2" uuid)" || die "UUID has not changed!" diff --git a/test/shell/pvcreate-bootloaderarea.sh b/test/shell/pvcreate-bootloaderarea.sh index 68c0882..80e1b3b 100644 --- a/test/shell/pvcreate-bootloaderarea.sh +++ b/test/shell/pvcreate-bootloaderarea.sh @@ -10,11 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Test pvcreate bootloader area support' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_devs 1 aux lvmconf 'global/suffix=0' 'global/units="b"' diff --git a/test/shell/pvcreate-ff.sh b/test/shell/pvcreate-ff.sh index 4ac5a61..2f620bd 100644 --- a/test/shell/pvcreate-ff.sh +++ b/test/shell/pvcreate-ff.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 2 pvcreate "$dev1" diff --git a/test/shell/pvcreate-metadata0.sh b/test/shell/pvcreate-metadata0.sh index 48ebc83..298edd0 100644 --- a/test/shell/pvcreate-metadata0.sh +++ b/test/shell/pvcreate-metadata0.sh @@ -15,9 +15,9 @@ # # 'Test pvcreate without metadata on all pvs' -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 2 128 diff --git a/test/shell/pvcreate-operation-md.sh b/test/shell/pvcreate-operation-md.sh index db53216..78e6cc5 100644 --- a/test/shell/pvcreate-operation-md.sh +++ b/test/shell/pvcreate-operation-md.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # skip this test if mdadm or sfdisk (or others) aren't available which sfdisk || skip diff --git a/test/shell/pvcreate-operation.sh b/test/shell/pvcreate-operation.sh index 6593877..3181722 100644 --- a/test/shell/pvcreate-operation.sh +++ b/test/shell/pvcreate-operation.sh @@ -8,9 +8,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux lvmconf 'devices/md_component_detection = 1' diff --git a/test/shell/pvcreate-usage.sh b/test/shell/pvcreate-usage.sh index ed3b4f4..5b74ed3 100644 --- a/test/shell/pvcreate-usage.sh +++ b/test/shell/pvcreate-usage.sh @@ -10,12 +10,11 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Test pvcreate option values' +SKIP_WITH_LVMPOLLD=1 PAGESIZE=$(getconf PAGESIZE) . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_devs 4 #COMM 'pvcreate rejects negative setphysicalvolumesize' diff --git a/test/shell/pvmove-cache-segtypes.sh b/test/shell/pvmove-cache-segtypes.sh index 2f09596..69554ef 100644 --- a/test/shell/pvmove-cache-segtypes.sh +++ b/test/shell/pvmove-cache-segtypes.sh @@ -10,6 +10,7 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description="ensure pvmove works with the cache segment types" +SKIP_WITH_CLVMD=1 . lib/inittest @@ -19,7 +20,6 @@ test_description="ensure pvmove works with the cache segment types" # This allows us to ensure that it is disallowed even when there are # stacking complications to consider. -test -e LOCAL_CLVMD && skip which md5sum || skip aux have_cache 1 3 0 || skip diff --git a/test/shell/pvmove-raid-segtypes.sh b/test/shell/pvmove-raid-segtypes.sh index 0accf4a..69cef2d 100644 --- a/test/shell/pvmove-raid-segtypes.sh +++ b/test/shell/pvmove-raid-segtypes.sh @@ -10,10 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description="ensure pvmove works with raid segment types" +SKIP_WITH_CLVMD=1 . lib/inittest -test -e LOCAL_CLVMD && skip which md5sum || skip aux have_raid 1 3 5 || skip diff --git a/test/shell/pvmove-resume-1.sh b/test/shell/pvmove-resume-1.sh index 0070826..5bd0e45 100644 --- a/test/shell/pvmove-resume-1.sh +++ b/test/shell/pvmove-resume-1.sh @@ -14,6 +14,8 @@ # 2 pvmove LVs in 2 VGs (1 per VG) +SKIP_WITH_CLVMD=1 + . lib/inittest aux prepare_pvs 4 30 @@ -245,8 +247,6 @@ pvmove_bg_single() { LVM_TEST_TAG="kill_me_$PREFIX" pvmove -b "$dev2" } -test -e LOCAL_CLVMD && skip - test_pvmove_resume lvchange_single test_pvmove_resume lvchange_all test_pvmove_resume vgchange_single diff --git a/test/shell/pvmove-resume-2.sh b/test/shell/pvmove-resume-2.sh index 7ae3721..ff3806f 100644 --- a/test/shell/pvmove-resume-2.sh +++ b/test/shell/pvmove-resume-2.sh @@ -14,6 +14,8 @@ # Moving 2 LVs in VG variant +SKIP_WITH_CLVMD=1 + . lib/inittest aux prepare_pvs 2 30 @@ -189,8 +191,6 @@ pvmove_bg_single() { LVM_TEST_TAG="kill_me_$PREFIX" pvmove -b "$dev1" } -test -e LOCAL_CLVMD && skip - test_pvmove_resume lvchange_single test_pvmove_resume lvchange_all test_pvmove_resume vgchange_single diff --git a/test/shell/pvmove-resume-multiseg.sh b/test/shell/pvmove-resume-multiseg.sh index 713534f..c113252 100644 --- a/test/shell/pvmove-resume-multiseg.sh +++ b/test/shell/pvmove-resume-multiseg.sh @@ -14,6 +14,8 @@ # Multisegment variant w/ 2 pvmoves LVs per VG +SKIP_WITH_CLVMD=1 + . lib/inittest aux prepare_pvs 5 30 @@ -219,8 +221,6 @@ pvmove_bg_single() { LVM_TEST_TAG="kill_me_$PREFIX" pvmove -b "$dev3" } -test -e LOCAL_CLVMD && skip - test_pvmove_resume lvchange_single test_pvmove_resume lvchange_all test_pvmove_resume vgchange_single diff --git a/test/shell/pvmove-thin-segtypes.sh b/test/shell/pvmove-thin-segtypes.sh index 33128d5..1d4e251 100644 --- a/test/shell/pvmove-thin-segtypes.sh +++ b/test/shell/pvmove-thin-segtypes.sh @@ -10,9 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description="ensure pvmove works with thin segment types" +SKIP_WITH_CLVMD=1 . lib/inittest -test -e LOCAL_CLVMD && skip + which md5sum || skip aux have_thin 1 8 0 || skip diff --git a/test/shell/pvremove-thin.sh b/test/shell/pvremove-thin.sh index 2400ac5..cb97bc2 100644 --- a/test/shell/pvremove-thin.sh +++ b/test/shell/pvremove-thin.sh @@ -12,9 +12,9 @@ # Checks we are not reading our own devices # https://bugzilla.redhat.com/show_bug.cgi?id=1064374 -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg diff --git a/test/shell/pvremove-usage.sh b/test/shell/pvremove-usage.sh index 5c55839..5ac79c2 100644 --- a/test/shell/pvremove-usage.sh +++ b/test/shell/pvremove-usage.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 3 pvcreate "$dev1" diff --git a/test/shell/pvremove-warnings.sh b/test/shell/pvremove-warnings.sh index f7ac95b..2361313 100644 --- a/test/shell/pvremove-warnings.sh +++ b/test/shell/pvremove-warnings.sh @@ -8,9 +8,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 2 pvcreate "$dev1" "$dev2" diff --git a/test/shell/read-ahead.sh b/test/shell/read-ahead.sh index 322a527..82a0707 100644 --- a/test/shell/read-ahead.sh +++ b/test/shell/read-ahead.sh @@ -14,11 +14,10 @@ # test_description='Test read-ahead functionality' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_vg 5 #COMM "test various read ahead settings (bz450922)" diff --git a/test/shell/select-report.sh b/test/shell/select-report.sh index b859199..20d57c7 100644 --- a/test/shell/select-report.sh +++ b/test/shell/select-report.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 6 16 diff --git a/test/shell/select-tools-thin.sh b/test/shell/select-tools-thin.sh index 4cabde1..7ea2852 100644 --- a/test/shell/select-tools-thin.sh +++ b/test/shell/select-tools-thin.sh @@ -9,12 +9,12 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux have_thin 1 0 0 || skip aux prepare_pvs 1 16 diff --git a/test/shell/select-tools.sh b/test/shell/select-tools.sh index 631c637..249d38b 100644 --- a/test/shell/select-tools.sh +++ b/test/shell/select-tools.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 4 12 diff --git a/test/shell/snapshot-autoumount-dmeventd.sh b/test/shell/snapshot-autoumount-dmeventd.sh index e82a757..0cd3a20 100644 --- a/test/shell/snapshot-autoumount-dmeventd.sh +++ b/test/shell/snapshot-autoumount-dmeventd.sh @@ -11,9 +11,9 @@ # no automatic extensions please -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest which mkfs.ext2 || skip diff --git a/test/shell/snapshot-cluster.sh b/test/shell/snapshot-cluster.sh index 47c8d2e..2a156f3 100644 --- a/test/shell/snapshot-cluster.sh +++ b/test/shell/snapshot-cluster.sh @@ -12,9 +12,9 @@ # Testing renaming snapshots in cluster # https://bugzilla.redhat.com/show_bug.cgi?id=1136925 -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 1 diff --git a/test/shell/snapshot-lvm1.sh b/test/shell/snapshot-lvm1.sh index 0abba0c..4986835 100644 --- a/test/shell/snapshot-lvm1.sh +++ b/test/shell/snapshot-lvm1.sh @@ -12,9 +12,9 @@ # regression test for lvmetad reporting error: # Internal error: LV snap_with_lvm1_meta (00000000000000000000000000000001) missing from preload metadata -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 2 diff --git a/test/shell/snapshot-maxsize.sh b/test/shell/snapshot-maxsize.sh index a96bd7f..5adf3f5 100644 --- a/test/shell/snapshot-maxsize.sh +++ b/test/shell/snapshot-maxsize.sh @@ -12,9 +12,9 @@ # Testing calculation of snapshot space # https://bugzilla.redhat.com/show_bug.cgi?id=1035871 -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 1 diff --git a/test/shell/snapshot-reactivate.sh b/test/shell/snapshot-reactivate.sh index 9e9b4b7..eb3d686 100644 --- a/test/shell/snapshot-reactivate.sh +++ b/test/shell/snapshot-reactivate.sh @@ -15,9 +15,10 @@ # is triggered by kernel 3.14-rc[1..5] # http://www.redhat.com/archives/dm-devel/2014-March/msg00005.html # -. lib/inittest -test -e LOCAL_LVMPOLLD && skip +SKIP_WITH_LVMPOLLD=1 + +. lib/inittest # Snapshot should remain unmodified check_s_() { diff --git a/test/shell/snapshot-remove-dmsetup.sh b/test/shell/snapshot-remove-dmsetup.sh index 952a7bb..395fad1 100644 --- a/test/shell/snapshot-remove-dmsetup.sh +++ b/test/shell/snapshot-remove-dmsetup.sh @@ -11,14 +11,14 @@ # check if 'dmsetup --noflush' will work properly for mounted snapshot +SKIP_WITH_CLVMD=1 +SKIP_WITH_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 + . lib/inittest which mkfs.ext2 || skip -test -e LOCAL_LVMETAD && skip -test -e LOCAL_LVMPOLLD && skip -test -e LOCAL_CLVMD && skip - aux prepare_vg 5 # Create stacked device diff --git a/test/shell/snapshot-rename.sh b/test/shell/snapshot-rename.sh index 47bf213..7450aec 100644 --- a/test/shell/snapshot-rename.sh +++ b/test/shell/snapshot-rename.sh @@ -12,9 +12,9 @@ # Testing renaming snapshots (had problem in cluster) # https://bugzilla.redhat.com/show_bug.cgi?id=1136925 -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 1 diff --git a/test/shell/snapshot-usage-exa.sh b/test/shell/snapshot-usage-exa.sh index 133014b..15634ba 100644 --- a/test/shell/snapshot-usage-exa.sh +++ b/test/shell/snapshot-usage-exa.sh @@ -12,9 +12,9 @@ # Check very large device size (upto 15Exa bytes) # this needs 64bit arch -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux can_use_16T || skip diff --git a/test/shell/snapshot-usage.sh b/test/shell/snapshot-usage.sh index d15eee1..1abb239 100644 --- a/test/shell/snapshot-usage.sh +++ b/test/shell/snapshot-usage.sh @@ -11,9 +11,9 @@ # no automatic extensions please -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest MKFS=mkfs.ext2 which $MKFS || skip diff --git a/test/shell/stray-device-node.sh b/test/shell/stray-device-node.sh index 09d9e75..a50015b 100644 --- a/test/shell/stray-device-node.sh +++ b/test/shell/stray-device-node.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 3 cp -r "$dev1" "$DM_DEV_DIR/stray" diff --git a/test/shell/system_id.sh b/test/shell/system_id.sh index 5bb5c3a..0218b95 100644 --- a/test/shell/system_id.sh +++ b/test/shell/system_id.sh @@ -10,11 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Test system_id' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_devs 1 # with clvm enabled, vgcreate with no -c option creates a clustered vg, diff --git a/test/shell/tags.sh b/test/shell/tags.sh index 37d738f..2823dc9 100644 --- a/test/shell/tags.sh +++ b/test/shell/tags.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 4 diff --git a/test/shell/test-partition.sh b/test/shell/test-partition.sh index 0cdaa87..b6b9b20 100644 --- a/test/shell/test-partition.sh +++ b/test/shell/test-partition.sh @@ -15,12 +15,12 @@ # +SKIP_WITH_LVMPOLLD=1 + LVM_TEST_CONFIG_DEVICES="types = [\"device-mapper\", 142]" . lib/inittest -test -e LOCAL_LVMPOLLD && skip - which sfdisk || skip aux prepare_pvs 1 30 diff --git a/test/shell/thin-autoumount-dmeventd.sh b/test/shell/thin-autoumount-dmeventd.sh index 5f1fc6b..9c5ce87 100644 --- a/test/shell/thin-autoumount-dmeventd.sh +++ b/test/shell/thin-autoumount-dmeventd.sh @@ -11,6 +11,8 @@ # no automatic extensions, just umount +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} is_dir_mounted_() @@ -20,8 +22,6 @@ is_dir_mounted_() . lib/inittest -test -e LOCAL_LVMPOLLD && skip - # # Main # diff --git a/test/shell/thin-defaults.sh b/test/shell/thin-defaults.sh index 3e49390..432193a 100644 --- a/test/shell/thin-defaults.sh +++ b/test/shell/thin-defaults.sh @@ -11,12 +11,12 @@ # test defaults entered through lvm.conf +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - # # Main # diff --git a/test/shell/thin-restore.sh b/test/shell/thin-restore.sh index 3516e2d..578e0a7 100644 --- a/test/shell/thin-restore.sh +++ b/test/shell/thin-restore.sh @@ -11,12 +11,12 @@ # test restore operation of thin pool metadata +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - # # Main # diff --git a/test/shell/thin-volume-list.sh b/test/shell/thin-volume-list.sh index c3dab25..9fc0ff2 100644 --- a/test/shell/thin-volume-list.sh +++ b/test/shell/thin-volume-list.sh @@ -11,12 +11,12 @@ # test pool behaviour when volume_list masks activation +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - # # Main # diff --git a/test/shell/topology-support.sh b/test/shell/topology-support.sh index ebb7220..2b16cf6 100644 --- a/test/shell/topology-support.sh +++ b/test/shell/topology-support.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest which mkfs.ext3 || skip diff --git a/test/shell/unknown-segment.sh b/test/shell/unknown-segment.sh index a1d3696..c2148ba 100644 --- a/test/shell/unknown-segment.sh +++ b/test/shell/unknown-segment.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 4 diff --git a/test/shell/unlost-pv.sh b/test/shell/unlost-pv.sh index 76bf8a1..298b2a8 100644 --- a/test/shell/unlost-pv.sh +++ b/test/shell/unlost-pv.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest check_() { # vgscan needs --cache option for direct scan if lvmetad is used diff --git a/test/shell/vg-name-from-env.sh b/test/shell/vg-name-from-env.sh index a3d1b9a..ac21772 100644 --- a/test/shell/vg-name-from-env.sh +++ b/test/shell/vg-name-from-env.sh @@ -6,11 +6,10 @@ # of the GNU General Public License v.2. test_description='Test the vg name for an lv from env var' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_devs 2 pvcreate "$dev1" diff --git a/test/shell/vgcfgbackup-usage.sh b/test/shell/vgcfgbackup-usage.sh index b0faf38..b5c2d40 100644 --- a/test/shell/vgcfgbackup-usage.sh +++ b/test/shell/vgcfgbackup-usage.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 4 diff --git a/test/shell/vgchange-many.sh b/test/shell/vgchange-many.sh index 7b3d557..7cd7e42 100644 --- a/test/shell/vgchange-many.sh +++ b/test/shell/vgchange-many.sh @@ -10,9 +10,9 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA # Check perfomance of activation and deactivation -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # FIXME: lvmetad fails with i.e. 1500 device on memory failure... diff --git a/test/shell/vgchange-maxlv.sh b/test/shell/vgchange-maxlv.sh index 613d985..886fa50 100644 --- a/test/shell/vgchange-maxlv.sh +++ b/test/shell/vgchange-maxlv.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_dmeventd aux prepare_pvs 3 diff --git a/test/shell/vgchange-partial.sh b/test/shell/vgchange-partial.sh index f9c3ac0..2dc3da7 100644 --- a/test/shell/vgchange-partial.sh +++ b/test/shell/vgchange-partial.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 2 diff --git a/test/shell/vgchange-sysinit.sh b/test/shell/vgchange-sysinit.sh index 8450d63..a4f078c 100644 --- a/test/shell/vgchange-sysinit.sh +++ b/test/shell/vgchange-sysinit.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_CLVMD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_CLVMD && skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest which mkfs.ext3 || skip diff --git a/test/shell/vgchange-usage.sh b/test/shell/vgchange-usage.sh index 077b447..370dba8 100644 --- a/test/shell/vgchange-usage.sh +++ b/test/shell/vgchange-usage.sh @@ -10,11 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Exercise some vgchange diagnostics' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_pvs 4 pvcreate --metadatacopies 0 "$dev1" diff --git a/test/shell/vgck.sh b/test/shell/vgck.sh index a4930aa..dfabac5 100644 --- a/test/shell/vgck.sh +++ b/test/shell/vgck.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 3 lvcreate -n blabla -L 1 $vg diff --git a/test/shell/vgcreate-many-pvs.sh b/test/shell/vgcreate-many-pvs.sh index bb78caf..c2f1ef3 100644 --- a/test/shell/vgcreate-many-pvs.sh +++ b/test/shell/vgcreate-many-pvs.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # # Test to exercise larger number of PVs in a VG diff --git a/test/shell/vgcreate-usage.sh b/test/shell/vgcreate-usage.sh index df0094c..04d0daa 100644 --- a/test/shell/vgcreate-usage.sh +++ b/test/shell/vgcreate-usage.sh @@ -10,11 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Exercise some vgcreate diagnostics' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_devs 3 pvcreate "$dev1" "$dev2" pvcreate --metadatacopies 0 "$dev3" diff --git a/test/shell/vgextend-restoremissing.sh b/test/shell/vgextend-restoremissing.sh index d8de96d..4bc2439 100644 --- a/test/shell/vgextend-restoremissing.sh +++ b/test/shell/vgextend-restoremissing.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 3 diff --git a/test/shell/vgextend-usage.sh b/test/shell/vgextend-usage.sh index 65665bc..e0c30cf 100644 --- a/test/shell/vgextend-usage.sh +++ b/test/shell/vgextend-usage.sh @@ -13,9 +13,9 @@ # Exercise various vgextend commands # -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 5 diff --git a/test/shell/vgimportclone.sh b/test/shell/vgimportclone.sh index 53c8b00..8589f05 100644 --- a/test/shell/vgimportclone.sh +++ b/test/shell/vgimportclone.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 2 diff --git a/test/shell/vgmerge-operation.sh b/test/shell/vgmerge-operation.sh index 42976c2..136e56d 100644 --- a/test/shell/vgmerge-operation.sh +++ b/test/shell/vgmerge-operation.sh @@ -10,11 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA test_description='Test vgmerge operation' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_pvs 4 64 # 'vgmerge succeeds with single linear LV in source VG' diff --git a/test/shell/vgmerge-usage.sh b/test/shell/vgmerge-usage.sh index 981a964..f344281 100644 --- a/test/shell/vgmerge-usage.sh +++ b/test/shell/vgmerge-usage.sh @@ -10,11 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA # 'Test vgmerge command options for validity' +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux prepare_pvs 4 # 'vgmerge normal operation' diff --git a/test/shell/vgreduce-removemissing-snapshot.sh b/test/shell/vgreduce-removemissing-snapshot.sh index 1bea397..1023159 100644 --- a/test/shell/vgreduce-removemissing-snapshot.sh +++ b/test/shell/vgreduce-removemissing-snapshot.sh @@ -9,9 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_CLVMD=1 +SKIP_WITH_LVMPOLLD=1 -exit 0 +. lib/inittest # # Snapshots of 'mirrors' are not supported. They can no longer be created. @@ -21,7 +22,7 @@ exit 0 aux prepare_vg 5 -lvcreate --type mirror -m 3 --ig -L 2M -n 4way $vg "$dev1" "$dev2" "$dev3" "$dev4" "$dev5":0 +lvcreate --type mirror -m 3 -L 2M -n 4way $vg "$dev1" "$dev2" "$dev3" "$dev4" "$dev5":0 lvcreate -s $vg/4way -L 2M -n snap lvcreate -i 2 -L 2M $vg "$dev1" "$dev2" -n stripe diff --git a/test/shell/vgreduce-usage.sh b/test/shell/vgreduce-usage.sh index 9f55395..a4faa16 100644 --- a/test/shell/vgreduce-usage.sh +++ b/test/shell/vgreduce-usage.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 4 diff --git a/test/shell/vgremove-corrupt-vg.sh b/test/shell/vgremove-corrupt-vg.sh index dc3c9db..59cb9b7 100644 --- a/test/shell/vgremove-corrupt-vg.sh +++ b/test/shell/vgremove-corrupt-vg.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 3 lvcreate -n blabla -L 1 $vg -an --zero n diff --git a/test/shell/vgrename-usage.sh b/test/shell/vgrename-usage.sh index 59c3397..92070e8 100644 --- a/test/shell/vgrename-usage.sh +++ b/test/shell/vgrename-usage.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 4 pvcreate "$dev1" "$dev2" diff --git a/test/shell/vgsplit-operation.sh b/test/shell/vgsplit-operation.sh index 975e873..963f4c1 100644 --- a/test/shell/vgsplit-operation.sh +++ b/test/shell/vgsplit-operation.sh @@ -12,12 +12,12 @@ # Test vgsplit operation, including different LV types # disable lvmetad logging as it bogs down test systems +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_LVMETAD_DEBUG_OPTS=${LVM_TEST_LVMETAD_DEBUG_OPTS-} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - COMM() { LAST_TEST="$@" } diff --git a/test/shell/vgsplit-raid.sh b/test/shell/vgsplit-raid.sh index 2fe734f..7da8eff 100644 --- a/test/shell/vgsplit-raid.sh +++ b/test/shell/vgsplit-raid.sh @@ -10,11 +10,10 @@ # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA # Test vgsplit operation, including different LV types +SKIP_WITH_LVMPOLLD=1 . lib/inittest -test -e LOCAL_LVMPOLLD && skip - COMM() { LAST_TEST="$@" } diff --git a/test/shell/vgsplit-stacked.sh b/test/shell/vgsplit-stacked.sh index 6a04f07..237a054 100644 --- a/test/shell/vgsplit-stacked.sh +++ b/test/shell/vgsplit-stacked.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux extend_filter_LVMTEST aux prepare_pvs 3 diff --git a/test/shell/vgsplit-thin.sh b/test/shell/vgsplit-thin.sh index 754f8ba..9d60cc7 100644 --- a/test/shell/vgsplit-thin.sh +++ b/test/shell/vgsplit-thin.sh @@ -11,12 +11,12 @@ # Test vgsplit command options for validity +SKIP_WITH_LVMPOLLD=1 + export LVM_TEST_THIN_REPAIR_CMD=${LVM_TEST_THIN_REPAIR_CMD-/bin/false} . lib/inittest -test -e LOCAL_LVMPOLLD && skip - aux have_thin 1 0 0 || skip aux prepare_devs 5 diff --git a/test/shell/vgsplit-usage.sh b/test/shell/vgsplit-usage.sh index b0422d5..2ca8dc7 100644 --- a/test/shell/vgsplit-usage.sh +++ b/test/shell/vgsplit-usage.sh @@ -11,9 +11,9 @@ # Test vgsplit command options for validity -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 5 diff --git a/test/shell/zero-usage.sh b/test/shell/zero-usage.sh index df90529..ee04aa1 100644 --- a/test/shell/zero-usage.sh +++ b/test/shell/zero-usage.sh @@ -11,9 +11,9 @@ # Basic usage of zero target -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest which md5sum || skip

1 0

master - tests: avoid reading utils when skipping
by Zdenek Kabelac 27 Oct '15

27 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=76cff10a734a7c1e… Commit: 76cff10a734a7c1e26b3835ff967dac0b7e46bcb Parent: 1af2ab10d02c2f02891d5047af563651fbe294bb Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 27 15:08:43 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 27 16:00:09 2015 +0100 tests: avoid reading utils when skipping Save even more CPU/time and avoid reading utils, when skipping test. --- test/lib/inittest.sh | 26 +++++++++++++++++--------- 1 files changed, 17 insertions(+), 9 deletions(-) diff --git a/test/lib/inittest.sh b/test/lib/inittest.sh index beb0210..6fde23a 100644 --- a/test/lib/inittest.sh +++ b/test/lib/inittest.sh @@ -1,5 +1,5 @@ #!/usr/bin/env bash -# Copyright (C) 2011-2012 Red Hat, Inc. All rights reserved. +# Copyright (C) 2011-2015 Red Hat, Inc. All rights reserved. # # This copyrighted material is made available to anyone wishing to use, # modify, copy, or redistribute it subject to the terms and conditions @@ -9,6 +9,11 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +initskip() { + test "$#" -eq 0 || echo "TEST SKIPPED: $@" + exit 200 +} + # sanitize the environment LANG=C LC_ALL=C @@ -20,19 +25,22 @@ TESTNAME=${0##*/} PS4='#${BASH_SOURCE[0]##*/}:${LINENO}+ ' export TESTNAME PS4 -unset CDPATH - -# grab some common utilities -. lib/utils - if test -n "$LVM_TEST_FLAVOUR"; then . lib/flavour-$LVM_TEST_FLAVOUR fi -test -n "$SKIP_WITHOUT_CLVMD" && test "$LVM_TEST_LOCKING" -ne 3 && skip -test -n "$SKIP_WITHOUT_LVMETAD" && test -z "$LVM_TEST_LVMETAD" && skip -test -n "$SKIP_WITH_LVMPOLLD" && test -n "$LVM_TEST_LVMPOLLD" && skip +test -n "$SKIP_WITHOUT_CLVMD" && test "$LVM_TEST_LOCKING" -ne 3 && initskip +test -n "$SKIP_WITH_CLVMD" && test "$LVM_TEST_LOCKING" -eq 3 && initskip +test -n "$SKIP_WITHOUT_LVMETAD" && test -z "$LVM_TEST_LVMETAD" && initskip +test -n "$SKIP_WITH_LVMETAD" && test -n "$LVM_TEST_LVMETAD" && initskip + +test -n "$SKIP_WITH_LVMPOLLD" && test -n "$LVM_TEST_LVMPOLLD" && initskip + +unset CDPATH + +# grab some common utilities +. lib/utils TESTOLDPWD=$(pwd) COMMON_PREFIX="LVMTEST"

1 0

master - cleanup: snapshots of snapshots message
by Zdenek Kabelac 27 Oct '15

27 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=1af2ab10d02c2f02… Commit: 1af2ab10d02c2f02891d5047af563651fbe294bb Parent: 729f489009b4dfc8488265bc64414edcfeae0755 Author: Marian Csontos <mcsontos(a)redhat.com> AuthorDate: Thu Oct 22 13:32:52 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 27 11:42:48 2015 +0100 cleanup: snapshots of snapshots message No plans to support thick snapshost of snapshots. --- WHATS_NEW | 1 + lib/metadata/lv_manip.c | 3 +-- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 264b3f7..869e506 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Dropping 'yet' from error of unsupported thick snapshot of snapshots. Do not support unpartitioned DASD devices with CDL formatted with pvcreate. For thins use flush for suspend only when volume size is reduced. Enable code which detects the need of flush during suspend. diff --git a/lib/metadata/lv_manip.c b/lib/metadata/lv_manip.c index 8029dab..8e20dfa 100644 --- a/lib/metadata/lv_manip.c +++ b/lib/metadata/lv_manip.c @@ -7248,8 +7248,7 @@ static struct logical_volume *_lv_create_an_lv(struct volume_group *vg, return NULL; } if (lv_is_cow(origin_lv)) { - log_error("Snapshots of snapshots are not " - "supported yet."); + log_error("Snapshots of snapshots are not supported."); return NULL; } if (lv_is_locked(origin_lv)) {

1 0

master - pvcreate: don't support unpartitioned DASD devices with CDL formatted
by Zdenek Kabelac 27 Oct '15

27 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=729f489009b4dfc8… Commit: 729f489009b4dfc8488265bc64414edcfeae0755 Parent: 5d76bdcdbdf6bef81872e8eef2cce156cfce228f Author: Lidong Zhong <lzhong(a)suse.com> AuthorDate: Mon Oct 12 15:09:11 2015 +0800 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 27 11:42:47 2015 +0100 pvcreate: don't support unpartitioned DASD devices with CDL formatted The former patch(dab3ebce4c7) is a little bit strict. For example, it is OK to create PV on unpartitioned DASD devices with LDL formatted. So after lvm version containing the patch, LVs created on those devices could not be found. Signed-off-by: Lidong Zhong <lzhong(a)suse.com> --- WHATS_NEW | 1 + lib/Makefile.in | 1 + lib/device/dev-dasd.c | 74 +++++++++++++++++++++++++++++++++++++++++++++++++ lib/device/dev-type.c | 2 +- lib/device/dev-type.h | 1 + 5 files changed, 78 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index bced219..264b3f7 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Do not support unpartitioned DASD devices with CDL formatted with pvcreate. For thins use flush for suspend only when volume size is reduced. Enable code which detects the need of flush during suspend. Ensure --use-policy will resize volume to fit below threshold. diff --git a/lib/Makefile.in b/lib/Makefile.in index 3e0e6f8..00014c2 100644 --- a/lib/Makefile.in +++ b/lib/Makefile.in @@ -62,6 +62,7 @@ SOURCES =\ device/dev-swap.c \ device/dev-type.c \ device/dev-luks.c \ + device/dev-dasd.c \ display/display.c \ error/errseg.c \ unknown/unknown.c \ diff --git a/lib/device/dev-dasd.c b/lib/device/dev-dasd.c new file mode 100644 index 0000000..28156ca --- /dev/null +++ b/lib/device/dev-dasd.c @@ -0,0 +1,74 @@ +/* + * Copyright (C) 2015 Red Hat, Inc. All rights reserved. + * + * This file is part of LVM2. + * + * This copyrighted material is made available to anyone wishing to use, + * modify, copy, or redistribute it subject to the terms and conditions + * of the GNU Lesser General Public License v.2.1. + * + * You should have received a copy of the GNU Lesser General Public License + * along with this program; if not, write to the Free Software Foundation, + * Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "lib.h" +#include "metadata.h" +#include "dev-type.h" +#include <sys/ioctl.h> + +typedef struct dasd_information_t { + unsigned int devno; /* S/390 devno */ + unsigned int real_devno; /* for aliases */ + unsigned int schid; /* S/390 subchannel identifier */ + unsigned int cu_type : 16; /* from SenseID */ + unsigned int cu_model : 8; /* from SenseID */ + unsigned int dev_type : 16; /* from SenseID */ + unsigned int dev_model : 8; /* from SenseID */ + unsigned int open_count; + unsigned int req_queue_len; + unsigned int chanq_len; /* length of chanq */ + char type[4]; /* from discipline.name, 'none' for unknown */ + unsigned int status; /* current device level */ + unsigned int label_block; /* where to find the VOLSER */ + unsigned int FBA_layout; /* fixed block size (like AIXVOL) */ + unsigned int characteristics_size; + unsigned int confdata_size; + unsigned char characteristics[64];/*from read_device_characteristics */ + unsigned char configuration_data[256];/*from read_configuration_data */ + unsigned int format; /* format info like formatted/cdl/ldl/... */ + unsigned int features; /* dasd features like 'ro',... */ + unsigned int reserved0; /* reserved for further use ,... */ + unsigned int reserved1; /* reserved for further use ,... */ + unsigned int reserved2; /* reserved for further use ,... */ + unsigned int reserved3; /* reserved for further use ,... */ + unsigned int reserved4; /* reserved for further use ,... */ + unsigned int reserved5; /* reserved for further use ,... */ + unsigned int reserved6; /* reserved for further use ,... */ + unsigned int reserved7; /* reserved for further use ,... */ +} dasd_information_t; + +#define DASD_FORMAT_CDL 2 +#define BIODASDINFO2 _IOR('D', 3, dasd_information_t) + +int dasd_is_cdl_formatted(struct device *dev) +{ + int ret = 0; + dasd_information_t dasd_info; + + if (!dev_open_readonly(dev)) { + stack; + return ret; + } + + if (ioctl(dev->fd, BIODASDINFO2, &dasd_info) != 0) + goto_out; + + if (dasd_info.format == DASD_FORMAT_CDL) + ret = 1; +out: + if (!dev_close(dev)) + stack; + + return ret; +} diff --git a/lib/device/dev-type.c b/lib/device/dev-type.c index 1d03024..a66ef95 100644 --- a/lib/device/dev-type.c +++ b/lib/device/dev-type.c @@ -363,7 +363,7 @@ static int _native_dev_is_partitioned(struct dev_types *dt, struct device *dev) return 0; /* Unpartitioned DASD devices are not supported. */ - if (MAJOR(dev->dev) == dt->dasd_major) + if ((MAJOR(dev->dev) == dt->dasd_major) && dasd_is_cdl_formatted(dev)) return 1; if (!dev_open_readonly_quiet(dev)) { diff --git a/lib/device/dev-type.h b/lib/device/dev-type.h index 2a49b4b..f1d52ca 100644 --- a/lib/device/dev-type.h +++ b/lib/device/dev-type.h @@ -59,6 +59,7 @@ int major_is_scsi_device(struct dev_types *dt, int major); int dev_is_md(struct device *dev, uint64_t *sb); int dev_is_swap(struct device *dev, uint64_t *signature); int dev_is_luks(struct device *dev, uint64_t *signature); +int dasd_is_cdl_formatted(struct device *dev); /* Signature wiping. */ #define TYPE_LVM1_MEMBER 0x001

1 0

master - dmeventd: event string parser handles empty field
by Zdenek Kabelac 27 Oct '15

27 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=5d76bdcdbdf6bef8… Commit: 5d76bdcdbdf6bef81872e8eef2cce156cfce228f Parent: 3b5939bbbb974913db3ff7ac363cc7f1154b894e Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 27 11:12:59 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 27 11:42:40 2015 +0100 dmeventd: event string parser handles empty field Improve event string parser to avoid unneeded alloc+free. Daemon talk function uses '-' to mark NULL/missing field. So restore the NULL pointer back on parser. This should have made old tools like 'dmevent_tool' work again. As now 'uuid' or 'dso' could become NULL and then be properly used in _want_registered_device() function. Since lvm2 always fill these parameters, this change should have no effect on lvm2. --- WHATS_NEW_DM | 1 + daemons/dmeventd/dmeventd.c | 50 +++++++++++++++++++++++++++--------------- 2 files changed, 33 insertions(+), 18 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index e09ab63..961aec5 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Fix/restore parsing of empty field '-' when processing dmeventd event. Enhance dm_tree_node_size_changed() to recognize size reduction. Support exit on idle for dmenventd (1 hour). Add support to allow unmonitor device from plugin itself. diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index d6accbf..42c724b 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -473,34 +473,48 @@ static int _pthread_create_smallstack(pthread_t *t, void *(*fun)(void *), void * /* * Fetch a string off src and duplicate it into *ptr. - * Pay attention to zero-length strings. + * Pay attention to zero-length and 'empty' strings ('-'). */ /* FIXME? move to libdevmapper to share with the client lib (need to make delimiter a parameter then) */ static int _fetch_string(char **ptr, char **src, const int delimiter) { - int ret = 0; + int ret = 1; char *p; size_t len; + *ptr = NULL; /* Empty field returns NULL pointer */ - if ((p = strchr(*src, delimiter))) - *p = 0; - - if ((*ptr = dm_strdup(*src))) { - if ((len = strlen(*ptr))) - *src += len; - else { - dm_free(*ptr); - *ptr = NULL; + if ((*src)[0] == '-') { + /* Could be empty field '-', handle without allocation */ + if ((*src)[1] == '\0') { + (*src)++; + goto out; + } else if ((*src)[1] == delimiter) { + (*src) += 2; + goto out; } - - (*src)++; - ret = 1; } - if (p) - *p = delimiter; - + if ((p = strchr(*src, delimiter))) { + if (*src < p) { + *p = 0; /* Temporary exit with \0 */ + if (!(*ptr = dm_strdup(*src))) { + log_error("Failed to fetch item %s.", *src); + ret = 0; /* Allocation fail */ + } + *p = delimiter; + *src = p; + } + (*src)++; /* Skip delmiter, next field */ + } else if ((len = strlen(*src))) { + /* No delimiter, item ends with '\0' */ + if (!(*ptr = dm_strdup(*src))) { + log_error("Failed to fetch last item %s.", *src); + ret = 0; /* Fail */ + } + *src += len + 1; + } +out: return ret; }

1 0

master - tests: use skip flags
by Zdenek Kabelac 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=3b5939bbbb974913… Commit: 3b5939bbbb974913db3ff7ac363cc7f1154b894e Parent: a2dd1f6e191de8d0ba1292e01338d480958c5a0a Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 27 00:45:37 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 27 00:45:37 2015 +0100 tests: use skip flags corrected usage of skip flags. --- test/shell/activate-minor.sh | 3 ++- test/shell/activate-missing-segment.sh | 4 ++-- test/shell/activate-missing.sh | 4 ++-- test/shell/activate-partial.sh | 4 ++-- test/shell/activation-skip.sh | 4 ++-- test/shell/clvmd-restart.sh | 5 +++-- test/shell/lvmetad-ambiguous.sh | 6 +++--- test/shell/lvmetad-autoshutdown.sh | 6 +++--- test/shell/lvmetad-client-filter.sh | 6 +++--- test/shell/lvmetad-disabled.sh | 6 +++--- test/shell/lvmetad-dump.sh | 6 +++--- test/shell/lvmetad-lvm1.sh | 6 +++--- test/shell/lvmetad-lvscan-cache.sh | 6 +++--- test/shell/lvmetad-no-cluster.sh | 3 +++ test/shell/lvmetad-override.sh | 6 +++--- test/shell/lvmetad-pvs.sh | 4 ++-- test/shell/lvmetad-pvscan-cache.sh | 6 +++--- test/shell/lvmetad-pvscan-filter.sh | 6 +++--- test/shell/lvmetad-pvscan-md.sh | 6 +++--- test/shell/lvmetad-pvscan-nomda-bg.sh | 6 +++--- test/shell/lvmetad-pvscan-nomda.sh | 6 +++--- test/shell/lvmetad-restart.sh | 6 +++--- test/shell/lvmetad-test.sh | 4 ++-- test/shell/lvmetad-warning.sh | 6 +++--- 24 files changed, 65 insertions(+), 60 deletions(-) diff --git a/test/shell/activate-minor.sh b/test/shell/activate-minor.sh index d7ae5f9..086a1ca 100644 --- a/test/shell/activate-minor.sh +++ b/test/shell/activate-minor.sh @@ -9,12 +9,13 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + . lib/inittest # Just skip this test if minor is already in use... dmsetup info | tee info egrep "^Major, minor: *[0-9]+, 123" info && skip -test -e LOCAL_LVMPOLLD && skip aux prepare_vg 2 lvcreate -a n --zero n -l 1 -n foo $vg diff --git a/test/shell/activate-missing-segment.sh b/test/shell/activate-missing-segment.sh index 988c9cb..05d6a36 100644 --- a/test/shell/activate-missing-segment.sh +++ b/test/shell/activate-missing-segment.sh @@ -16,9 +16,9 @@ # instead lvconvert --repair them?) # - linear LVs with bits missing are not activated -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 2 diff --git a/test/shell/activate-missing.sh b/test/shell/activate-missing.sh index e33a6ef..3078b9d 100644 --- a/test/shell/activate-missing.sh +++ b/test/shell/activate-missing.sh @@ -16,9 +16,9 @@ # instead lvconvert --repair them?) # - linear LVs with bits missing are not activated -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 4 diff --git a/test/shell/activate-partial.sh b/test/shell/activate-partial.sh index 3720e1d..75a66e2 100644 --- a/test/shell/activate-partial.sh +++ b/test/shell/activate-partial.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 3 diff --git a/test/shell/activation-skip.sh b/test/shell/activation-skip.sh index 5259d9d..18bb218 100644 --- a/test/shell/activation-skip.sh +++ b/test/shell/activation-skip.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # Test skip activation flag -k|--setactivationskip diff --git a/test/shell/clvmd-restart.sh b/test/shell/clvmd-restart.sh index afe4834..5d8525f 100644 --- a/test/shell/clvmd-restart.sh +++ b/test/shell/clvmd-restart.sh @@ -13,11 +13,12 @@ export LVM_CLVMD_BINARY=clvmd export LVM_BINARY=lvm +SKIP_WITHOUT_CLVMD=1 +SKIP_WITH_LVMPOLLD=1 + . lib/inittest # only clvmd based test, skip otherwise -test -e LOCAL_LVMPOLLD && skip -test -e LOCAL_CLVMD || skip read LOCAL_CLVMD < LOCAL_CLVMD # TODO read from build, for now hardcoded diff --git a/test/shell/lvmetad-ambiguous.sh b/test/shell/lvmetad-ambiguous.sh index 777a63e..1711f71 100644 --- a/test/shell/lvmetad-ambiguous.sh +++ b/test/shell/lvmetad-ambiguous.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-autoshutdown.sh b/test/shell/lvmetad-autoshutdown.sh index fdee4a6..425ee33 100644 --- a/test/shell/lvmetad-autoshutdown.sh +++ b/test/shell/lvmetad-autoshutdown.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest kill -0 $(< LOCAL_LVMETAD) || die "lvmetad is already dead" diff --git a/test/shell/lvmetad-client-filter.sh b/test/shell/lvmetad-client-filter.sh index 5e6d967..4119e64 100644 --- a/test/shell/lvmetad-client-filter.sh +++ b/test/shell/lvmetad-client-filter.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-disabled.sh b/test/shell/lvmetad-disabled.sh index 9c71717..4f15165 100644 --- a/test/shell/lvmetad-disabled.sh +++ b/test/shell/lvmetad-disabled.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest kill $(< LOCAL_LVMETAD) while test -e "$TESTDIR/lvmetad.socket"; do echo -n .; sleep .1; done # wait for the socket close diff --git a/test/shell/lvmetad-dump.sh b/test/shell/lvmetad-dump.sh index 631636e..c9a7736 100644 --- a/test/shell/lvmetad-dump.sh +++ b/test/shell/lvmetad-dump.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 vgcreate $vg1 "$dev1" "$dev2" diff --git a/test/shell/lvmetad-lvm1.sh b/test/shell/lvmetad-lvm1.sh index f234b15..3b14fdd 100644 --- a/test/shell/lvmetad-lvm1.sh +++ b/test/shell/lvmetad-lvm1.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 2 pvcreate --metadatatype 1 "$dev1" diff --git a/test/shell/lvmetad-lvscan-cache.sh b/test/shell/lvmetad-lvscan-cache.sh index 220e464..a4a0982 100644 --- a/test/shell/lvmetad-lvscan-cache.sh +++ b/test/shell/lvmetad-lvscan-cache.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-no-cluster.sh b/test/shell/lvmetad-no-cluster.sh index 83fee47..0421c59 100644 --- a/test/shell/lvmetad-no-cluster.sh +++ b/test/shell/lvmetad-no-cluster.sh @@ -9,6 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITHOUT_CLVMD=1 +SKIP_WITH_LVMPOLLD=1 + . lib/inittest test -e LOCAL_CLVMD || skip diff --git a/test/shell/lvmetad-override.sh b/test/shell/lvmetad-override.sh index 15ada7b..a0a983b 100644 --- a/test/shell/lvmetad-override.sh +++ b/test/shell/lvmetad-override.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-pvs.sh b/test/shell/lvmetad-pvs.sh index 81174f4..3014a67 100644 --- a/test/shell/lvmetad-pvs.sh +++ b/test/shell/lvmetad-pvs.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 1 20000 pvs $(cat DEVICES) | grep "$dev1" diff --git a/test/shell/lvmetad-pvscan-cache.sh b/test/shell/lvmetad-pvscan-cache.sh index 33b4d94..f2a3afa 100644 --- a/test/shell/lvmetad-pvscan-cache.sh +++ b/test/shell/lvmetad-pvscan-cache.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-pvscan-filter.sh b/test/shell/lvmetad-pvscan-filter.sh index b7bdf33..8ec31ed 100644 --- a/test/shell/lvmetad-pvscan-filter.sh +++ b/test/shell/lvmetad-pvscan-filter.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-pvscan-md.sh b/test/shell/lvmetad-pvscan-md.sh index 9c5e3fa..82831e5 100644 --- a/test/shell/lvmetad-pvscan-md.sh +++ b/test/shell/lvmetad-pvscan-md.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest test -f /proc/mdstat && grep -q raid0 /proc/mdstat || \ modprobe raid0 || skip diff --git a/test/shell/lvmetad-pvscan-nomda-bg.sh b/test/shell/lvmetad-pvscan-nomda-bg.sh index e17dfd2..5cf4602 100644 --- a/test/shell/lvmetad-pvscan-nomda-bg.sh +++ b/test/shell/lvmetad-pvscan-nomda-bg.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest kill $(< LOCAL_LVMETAD) rm LOCAL_LVMETAD diff --git a/test/shell/lvmetad-pvscan-nomda.sh b/test/shell/lvmetad-pvscan-nomda.sh index f7e13a7..6a78b6c 100644 --- a/test/shell/lvmetad-pvscan-nomda.sh +++ b/test/shell/lvmetad-pvscan-nomda.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest kill $(< LOCAL_LVMETAD) rm LOCAL_LVMETAD diff --git a/test/shell/lvmetad-restart.sh b/test/shell/lvmetad-restart.sh index 8805bc8..18b2b2c 100644 --- a/test/shell/lvmetad-restart.sh +++ b/test/shell/lvmetad-restart.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-test.sh b/test/shell/lvmetad-test.sh index 146a7f1..7d47979 100644 --- a/test/shell/lvmetad-test.sh +++ b/test/shell/lvmetad-test.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-warning.sh b/test/shell/lvmetad-warning.sh index 44edb12..4c8cc59 100644 --- a/test/shell/lvmetad-warning.sh +++ b/test/shell/lvmetad-warning.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2

1 0

master - tests: use skip flags
by Zdenek Kabelac 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=46395f22ea6089a5… Commit: 46395f22ea6089a508d30e0ef2e85f0e036359da Parent: a2dd1f6e191de8d0ba1292e01338d480958c5a0a Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 26 20:09:58 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Mon Oct 26 23:57:48 2015 +0100 tests: use skip flags --- test/shell/activate-minor.sh | 3 ++- test/shell/activate-missing-segment.sh | 4 ++-- test/shell/activate-missing.sh | 4 ++-- test/shell/activate-partial.sh | 4 ++-- test/shell/activation-skip.sh | 4 ++-- test/shell/clvmd-restart.sh | 5 +++-- test/shell/lvmetad-ambiguous.sh | 6 +++--- test/shell/lvmetad-autoshutdown.sh | 6 +++--- test/shell/lvmetad-client-filter.sh | 6 +++--- test/shell/lvmetad-disabled.sh | 6 +++--- test/shell/lvmetad-dump.sh | 6 +++--- test/shell/lvmetad-lvm1.sh | 6 +++--- test/shell/lvmetad-lvscan-cache.sh | 6 +++--- test/shell/lvmetad-no-cluster.sh | 3 +++ test/shell/lvmetad-override.sh | 6 +++--- test/shell/lvmetad-pvs.sh | 4 ++-- test/shell/lvmetad-pvscan-cache.sh | 6 +++--- test/shell/lvmetad-pvscan-filter.sh | 6 +++--- test/shell/lvmetad-pvscan-md.sh | 6 +++--- test/shell/lvmetad-pvscan-nomda-bg.sh | 6 +++--- test/shell/lvmetad-pvscan-nomda.sh | 6 +++--- test/shell/lvmetad-restart.sh | 6 +++--- test/shell/lvmetad-test.sh | 4 ++-- test/shell/lvmetad-warning.sh | 6 +++--- 24 files changed, 65 insertions(+), 60 deletions(-) diff --git a/test/shell/activate-minor.sh b/test/shell/activate-minor.sh index d7ae5f9..086a1ca 100644 --- a/test/shell/activate-minor.sh +++ b/test/shell/activate-minor.sh @@ -9,12 +9,13 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITH_LVMPOLLD=1 + . lib/inittest # Just skip this test if minor is already in use... dmsetup info | tee info egrep "^Major, minor: *[0-9]+, 123" info && skip -test -e LOCAL_LVMPOLLD && skip aux prepare_vg 2 lvcreate -a n --zero n -l 1 -n foo $vg diff --git a/test/shell/activate-missing-segment.sh b/test/shell/activate-missing-segment.sh index 988c9cb..05d6a36 100644 --- a/test/shell/activate-missing-segment.sh +++ b/test/shell/activate-missing-segment.sh @@ -16,9 +16,9 @@ # instead lvconvert --repair them?) # - linear LVs with bits missing are not activated -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 2 diff --git a/test/shell/activate-missing.sh b/test/shell/activate-missing.sh index e33a6ef..3078b9d 100644 --- a/test/shell/activate-missing.sh +++ b/test/shell/activate-missing.sh @@ -16,9 +16,9 @@ # instead lvconvert --repair them?) # - linear LVs with bits missing are not activated -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 4 diff --git a/test/shell/activate-partial.sh b/test/shell/activate-partial.sh index 3720e1d..75a66e2 100644 --- a/test/shell/activate-partial.sh +++ b/test/shell/activate-partial.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_vg 3 diff --git a/test/shell/activation-skip.sh b/test/shell/activation-skip.sh index 5259d9d..18bb218 100644 --- a/test/shell/activation-skip.sh +++ b/test/shell/activation-skip.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest # Test skip activation flag -k|--setactivationskip diff --git a/test/shell/clvmd-restart.sh b/test/shell/clvmd-restart.sh index afe4834..5d8525f 100644 --- a/test/shell/clvmd-restart.sh +++ b/test/shell/clvmd-restart.sh @@ -13,11 +13,12 @@ export LVM_CLVMD_BINARY=clvmd export LVM_BINARY=lvm +SKIP_WITHOUT_CLVMD=1 +SKIP_WITH_LVMPOLLD=1 + . lib/inittest # only clvmd based test, skip otherwise -test -e LOCAL_LVMPOLLD && skip -test -e LOCAL_CLVMD || skip read LOCAL_CLVMD < LOCAL_CLVMD # TODO read from build, for now hardcoded diff --git a/test/shell/lvmetad-ambiguous.sh b/test/shell/lvmetad-ambiguous.sh index 777a63e..1711f71 100644 --- a/test/shell/lvmetad-ambiguous.sh +++ b/test/shell/lvmetad-ambiguous.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-autoshutdown.sh b/test/shell/lvmetad-autoshutdown.sh index fdee4a6..425ee33 100644 --- a/test/shell/lvmetad-autoshutdown.sh +++ b/test/shell/lvmetad-autoshutdown.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest kill -0 $(< LOCAL_LVMETAD) || die "lvmetad is already dead" diff --git a/test/shell/lvmetad-client-filter.sh b/test/shell/lvmetad-client-filter.sh index 5e6d967..4119e64 100644 --- a/test/shell/lvmetad-client-filter.sh +++ b/test/shell/lvmetad-client-filter.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-disabled.sh b/test/shell/lvmetad-disabled.sh index 9c71717..4f15165 100644 --- a/test/shell/lvmetad-disabled.sh +++ b/test/shell/lvmetad-disabled.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest kill $(< LOCAL_LVMETAD) while test -e "$TESTDIR/lvmetad.socket"; do echo -n .; sleep .1; done # wait for the socket close diff --git a/test/shell/lvmetad-dump.sh b/test/shell/lvmetad-dump.sh index 631636e..c9a7736 100644 --- a/test/shell/lvmetad-dump.sh +++ b/test/shell/lvmetad-dump.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 vgcreate $vg1 "$dev1" "$dev2" diff --git a/test/shell/lvmetad-lvm1.sh b/test/shell/lvmetad-lvm1.sh index f234b15..3b14fdd 100644 --- a/test/shell/lvmetad-lvm1.sh +++ b/test/shell/lvmetad-lvm1.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_devs 2 pvcreate --metadatatype 1 "$dev1" diff --git a/test/shell/lvmetad-lvscan-cache.sh b/test/shell/lvmetad-lvscan-cache.sh index 220e464..a4a0982 100644 --- a/test/shell/lvmetad-lvscan-cache.sh +++ b/test/shell/lvmetad-lvscan-cache.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-no-cluster.sh b/test/shell/lvmetad-no-cluster.sh index 83fee47..0421c59 100644 --- a/test/shell/lvmetad-no-cluster.sh +++ b/test/shell/lvmetad-no-cluster.sh @@ -9,6 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +SKIP_WITHOUT_CLVMD=1 +SKIP_WITH_LVMPOLLD=1 + . lib/inittest test -e LOCAL_CLVMD || skip diff --git a/test/shell/lvmetad-override.sh b/test/shell/lvmetad-override.sh index 15ada7b..a0a983b 100644 --- a/test/shell/lvmetad-override.sh +++ b/test/shell/lvmetad-override.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-pvs.sh b/test/shell/lvmetad-pvs.sh index 81174f4..21b404b 100644 --- a/test/shell/lvmetad-pvs.sh +++ b/test/shell/lvmetad-pvs.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 1 20000 pvs $(cat DEVICES) | grep "$dev1" diff --git a/test/shell/lvmetad-pvscan-cache.sh b/test/shell/lvmetad-pvscan-cache.sh index 33b4d94..f2a3afa 100644 --- a/test/shell/lvmetad-pvscan-cache.sh +++ b/test/shell/lvmetad-pvscan-cache.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-pvscan-filter.sh b/test/shell/lvmetad-pvscan-filter.sh index b7bdf33..8ec31ed 100644 --- a/test/shell/lvmetad-pvscan-filter.sh +++ b/test/shell/lvmetad-pvscan-filter.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-pvscan-md.sh b/test/shell/lvmetad-pvscan-md.sh index 9c5e3fa..82831e5 100644 --- a/test/shell/lvmetad-pvscan-md.sh +++ b/test/shell/lvmetad-pvscan-md.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest test -f /proc/mdstat && grep -q raid0 /proc/mdstat || \ modprobe raid0 || skip diff --git a/test/shell/lvmetad-pvscan-nomda-bg.sh b/test/shell/lvmetad-pvscan-nomda-bg.sh index e17dfd2..5cf4602 100644 --- a/test/shell/lvmetad-pvscan-nomda-bg.sh +++ b/test/shell/lvmetad-pvscan-nomda-bg.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest kill $(< LOCAL_LVMETAD) rm LOCAL_LVMETAD diff --git a/test/shell/lvmetad-pvscan-nomda.sh b/test/shell/lvmetad-pvscan-nomda.sh index f7e13a7..6a78b6c 100644 --- a/test/shell/lvmetad-pvscan-nomda.sh +++ b/test/shell/lvmetad-pvscan-nomda.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest kill $(< LOCAL_LVMETAD) rm LOCAL_LVMETAD diff --git a/test/shell/lvmetad-restart.sh b/test/shell/lvmetad-restart.sh index 8805bc8..18b2b2c 100644 --- a/test/shell/lvmetad-restart.sh +++ b/test/shell/lvmetad-restart.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-test.sh b/test/shell/lvmetad-test.sh index 146a7f1..7d47979 100644 --- a/test/shell/lvmetad-test.sh +++ b/test/shell/lvmetad-test.sh @@ -9,9 +9,9 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2 diff --git a/test/shell/lvmetad-warning.sh b/test/shell/lvmetad-warning.sh index 44edb12..4c8cc59 100644 --- a/test/shell/lvmetad-warning.sh +++ b/test/shell/lvmetad-warning.sh @@ -9,10 +9,10 @@ # along with this program; if not, write to the Free Software Foundation, # Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -. lib/inittest +SKIP_WITHOUT_LVMETAD=1 +SKIP_WITH_LVMPOLLD=1 -test -e LOCAL_LVMETAD || skip -test -e LOCAL_LVMPOLLD && skip +. lib/inittest aux prepare_pvs 2

1 0

master - tests: support skip flags
by Zdenek Kabelac 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=a2dd1f6e191de8d0… Commit: a2dd1f6e191de8d0ba1292e01338d480958c5a0a Parent: c301cc5d3840e1badd015164e0dcabfacc8f9599 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 26 20:09:47 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Mon Oct 26 23:57:47 2015 +0100 tests: support skip flags --- test/lib/inittest.sh | 13 +++++++++---- 1 files changed, 9 insertions(+), 4 deletions(-) diff --git a/test/lib/inittest.sh b/test/lib/inittest.sh index 28bc4e8..beb0210 100644 --- a/test/lib/inittest.sh +++ b/test/lib/inittest.sh @@ -25,6 +25,15 @@ unset CDPATH # grab some common utilities . lib/utils +if test -n "$LVM_TEST_FLAVOUR"; then + . lib/flavour-$LVM_TEST_FLAVOUR +fi + +test -n "$SKIP_WITHOUT_CLVMD" && test "$LVM_TEST_LOCKING" -ne 3 && skip +test -n "$SKIP_WITHOUT_LVMETAD" && test -z "$LVM_TEST_LVMETAD" && skip +test -n "$SKIP_WITH_LVMPOLLD" && test -n "$LVM_TEST_LVMPOLLD" && skip + + TESTOLDPWD=$(pwd) COMMON_PREFIX="LVMTEST" PREFIX="${COMMON_PREFIX}$$" @@ -54,10 +63,6 @@ test -n "$abs_top_builddir" && \ find "$TESTOLDPWD/lib" ! $ -name '*.sh' -o -name '*.[cdo]' \ -o -name '*~' $ -exec ln -s -t lib "{}" + -if test -n "$LVM_TEST_FLAVOUR"; then - . lib/flavour-$LVM_TEST_FLAVOUR -fi - DM_DEFAULT_NAME_MANGLING_MODE=none DM_DEV_DIR="$TESTDIR/dev" LVM_SYSTEM_DIR="$TESTDIR/etc"

1 0

master - tests: extend timer for 4 hours
by Zdenek Kabelac 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=c301cc5d3840e1ba… Commit: c301cc5d3840e1badd015164e0dcabfacc8f9599 Parent: ba41ee1dc94264f7ac8e61f8b1d56b10225b0d2f Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 26 11:04:58 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Mon Oct 26 23:57:47 2015 +0100 tests: extend timer for 4 hours Extend max time for test suite to 4 hours. Also replace some 'non-ascii' chars from source files and keep them plain ascii. --- aclocal.m4 | 2 +- test/lib/brick-shelltest.h | 9 ++++++--- 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/aclocal.m4 b/aclocal.m4 index 05e0ad5..67b2d59 100644 --- a/aclocal.m4 +++ b/aclocal.m4 @@ -15,7 +15,7 @@ m4_ifndef([AC_CONFIG_MACRO_DIRS], [m4_defun([_AM_CONFIG_MACRO_DIRS], [])m4_defun # pkg.m4 - Macros to locate and utilise pkg-config. -*- Autoconf -*- # serial 1 (pkg-config-0.24) # -# Copyright © 2004 Scott James Remnant <scott(a)netsplit.com>. +# Copyright (c) 2004 Scott James Remnant <scott(a)netsplit.com>. # # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by diff --git a/test/lib/brick-shelltest.h b/test/lib/brick-shelltest.h index 0659d8e..ccb4fb4 100644 --- a/test/lib/brick-shelltest.h +++ b/test/lib/brick-shelltest.h @@ -17,7 +17,7 @@ */ /* - * (c) 2014 Petr Ročkai <me(a)mornfall.net> + * (c) 2014 Petr Rockai <me(a)mornfall.net> * (c) 2014 Red Hat, Inc. */ @@ -81,6 +81,9 @@ #include "configure.h" +/* Timeout for the whole test suite in hours */ +static const unsigned TEST_SUITE_TIMEOUT = 4; + #ifndef BRICK_SHELLTEST_H #define BRICK_SHELLTEST_H @@ -1034,8 +1037,8 @@ struct Main { die = 1; } - if ( time(0) - start > 3 * 3600 ) { - std::cerr << "3 hours passed, giving up..." << std::endl; + if ( time(0) - start > (TEST_SUITE_TIMEOUT * 3600) ) { + std::cerr << TEST_SUITE_TIMEOUT << " hours passed, giving up..." << std::endl; die = 1; }

1 0

master - thin: limit no-flush using only for thin-pool
by Zdenek Kabelac 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=ba41ee1dc94264f7… Commit: ba41ee1dc94264f7ac8e61f8b1d56b10225b0d2f Parent: b702d677477ff6b37678fe2b6c38969792c09359 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 26 21:46:54 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Mon Oct 26 23:57:31 2015 +0100 thin: limit no-flush using only for thin-pool For this release keep usage of 'noflush' only for thin-volume/pool. For rest of keep - keep usage of 'noflush' flag purely for non-resized mirrors. --- WHATS_NEW | 2 +- lib/activate/dev_manager.c | 7 +++++++ 2 files changed, 8 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 80a198a..bced219 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,6 +1,6 @@ Version 2.02.133 - ====================================== - Enforce flush for suspend only when volume size is reduced. + For thins use flush for suspend only when volume size is reduced. Enable code which detects the need of flush during suspend. Ensure --use-policy will resize volume to fit below threshold. Correct percentage evaluation when checking thin-pool over threshold. diff --git a/lib/activate/dev_manager.c b/lib/activate/dev_manager.c index dc6d2fd..964eef5 100644 --- a/lib/activate/dev_manager.c +++ b/lib/activate/dev_manager.c @@ -3299,6 +3299,13 @@ static int _tree_action(struct dev_manager *dm, const struct logical_volume *lv, if ((dm_tree_node_size_changed(root) < 0)) dm->flush_required = 1; + /* Currently keep the code require flush for any + * non 'thin pool/volume, mirror' or with any size change */ + if (!lv_is_thin_volume(lv) && + !lv_is_thin_pool(lv) && + (!lv_is_mirror(lv) || dm_tree_node_size_changed(root))) + dm->flush_required = 1; + if (action == ACTIVATE) { if (!dm_tree_activate_children(root, dlid, DLID_SIZE)) goto_out;

1 0

master - dmevent: fix referencing
by Zdenek Kabelac 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=b702d677477ff6b3… Commit: b702d677477ff6b37678fe2b6c38969792c09359 Parent: 44ba862674c995e2ed297714d3aa2b2dbd731ad1 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 26 23:29:31 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Mon Oct 26 23:31:23 2015 +0100 dmevent: fix referencing Plugin increments DSO refcounter in _alloc_thread_status(). --- daemons/dmeventd/dmeventd.c | 1 - 1 files changed, 0 insertions(+), 1 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index b5f1be3..d6accbf 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -296,7 +296,6 @@ static struct dso_data *_lookup_dso(struct message_data *data) dm_list_iterate_items(dso_data, &_dso_registry) if (!strcmp(data->dso_name, dso_data->dso_name)) { - _lib_get(dso_data); ret = dso_data; break; }

1 0

master - toollib: fix wrong paren placement
by David Teigland 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=44ba862674c995e2… Commit: 44ba862674c995e2ed297714d3aa2b2dbd731ad1 Parent: 6624833839645cbe75af073fd1dd9ab8c78e0d03 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Mon Oct 26 16:27:26 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Mon Oct 26 16:27:26 2015 -0500 toollib: fix wrong paren placement --- tools/toollib.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/tools/toollib.c b/tools/toollib.c index 843a47b..59ba29f 100644 --- a/tools/toollib.c +++ b/tools/toollib.c @@ -3119,7 +3119,7 @@ int process_each_pv(struct cmd_context *cmd, return ret; } - if ((ret = _get_arg_devices(cmd, &arg_pvnames, &arg_devices) != ECMD_PROCESSED)) + if ((ret = _get_arg_devices(cmd, &arg_pvnames, &arg_devices)) != ECMD_PROCESSED) /* get_arg_devices reports the error for any PV names not found. */ ret_max = ECMD_FAILED;

1 0

master - pvs: fix missing PVs when VG is removed
by David Teigland 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=6624833839645cbe… Commit: 6624833839645cbe75af073fd1dd9ab8c78e0d03 Parent: b29593378f41f6dd401fb9582fba1a100f33dbbb Author: David Teigland <teigland(a)redhat.com> AuthorDate: Fri Oct 23 15:09:20 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Mon Oct 26 16:07:12 2015 -0500 pvs: fix missing PVs when VG is removed PVs could be missing from the 'pvs' output if their VG was removed at the same time that the 'pvs' command was run. To fix this: 1. If a VG is not found when processed, don't silently skip the PVs in it, as is done when the "skip" variable is set. 2. Repeat the VG search if some PVs are not found on the first search through all VGs. The second search uses a specific list of PVs that were missed the first time. testing: /dev/sdb is a PV /dev/sdd is a PV /dev/sdg is not a PV each test begins with: vgcreate test /dev/sdb /dev/sdd variations to test: vgremove -f test & pvs vgremove -f test & pvs -a vgremove -f test & pvs /dev/sdb /dev/sdd vgremove -f test & pvs /dev/sdg vgremove -f test & pvs /dev/sdb /dev/sdg The pvs command should always display /dev/sdb and /dev/sdd, either as a part of VG test or not. The pvs command should always print an error indicating that /dev/sdg could not be found. --- tools/toollib.c | 125 +++++++++++++++++++++++++++++++++++++++++++++++++++---- 1 files changed, 116 insertions(+), 9 deletions(-) diff --git a/tools/toollib.c b/tools/toollib.c index fba075a..843a47b 100644 --- a/tools/toollib.c +++ b/tools/toollib.c @@ -184,14 +184,16 @@ const char *skip_dev_dir(struct cmd_context *cmd, const char *vg_name, * If *skip is 1, it's OK for the caller to read the list of PVs in the VG. */ static int _ignore_vg(struct volume_group *vg, const char *vg_name, - struct dm_list *arg_vgnames, uint32_t read_flags, int *skip) + struct dm_list *arg_vgnames, uint32_t read_flags, + int *skip, int *notfound) { uint32_t read_error = vg_read_error(vg); + *skip = 0; + *notfound = 0; if ((read_error & FAILED_NOTFOUND) && (read_flags & READ_OK_NOTFOUND)) { - read_error &= ~FAILED_NOTFOUND; - *skip = 1; + *notfound = 1; return 0; } @@ -1924,6 +1926,7 @@ static int _process_vgnameid_list(struct cmd_context *cmd, uint32_t flags, int ret_max = ECMD_PROCESSED; int ret; int skip; + int notfound; int process_all = 0; /* @@ -1942,6 +1945,7 @@ static int _process_vgnameid_list(struct cmd_context *cmd, uint32_t flags, vg_name = vgnl->vg_name; vg_uuid = vgnl->vgid; skip = 0; + notfound = 0; if (!lockd_vg(cmd, vg_name, NULL, 0, &lockd_state)) { ret_max = ECMD_FAILED; @@ -1949,12 +1953,12 @@ static int _process_vgnameid_list(struct cmd_context *cmd, uint32_t flags, } vg = vg_read(cmd, vg_name, vg_uuid, flags, lockd_state); - if (_ignore_vg(vg, vg_name, arg_vgnames, flags, &skip)) { + if (_ignore_vg(vg, vg_name, arg_vgnames, flags, &skip, &notfound)) { stack; ret_max = ECMD_FAILED; goto endvg; } - if (skip) + if (skip || notfound) goto endvg; /* Process this VG? */ @@ -2392,6 +2396,7 @@ static int _process_lv_vgnameid_list(struct cmd_context *cmd, uint32_t flags, int ret_max = ECMD_PROCESSED; int ret; int skip; + int notfound; dm_list_iterate_items(vgnl, vgnameids_to_process) { if (sigint_caught()) @@ -2400,6 +2405,7 @@ static int _process_lv_vgnameid_list(struct cmd_context *cmd, uint32_t flags, vg_name = vgnl->vg_name; vg_uuid = vgnl->vgid; skip = 0; + notfound = 0; /* * arg_lvnames contains some elements that are just "vgname" @@ -2437,13 +2443,12 @@ static int _process_lv_vgnameid_list(struct cmd_context *cmd, uint32_t flags, } vg = vg_read(cmd, vg_name, vg_uuid, flags, lockd_state); - if (_ignore_vg(vg, vg_name, arg_vgnames, flags, &skip)) { + if (_ignore_vg(vg, vg_name, arg_vgnames, flags, &skip, &notfound)) { stack; ret_max = ECMD_FAILED; goto endvg; - } - if (skip) + if (skip || notfound) goto endvg; ret = process_each_lv_in_vg(cmd, vg, &lvnames, tags_arg, 0, @@ -2688,6 +2693,59 @@ static struct device_id_list *_device_list_find_pvid(struct dm_list *devices, st return NULL; } +static int _device_list_copy(struct cmd_context *cmd, struct dm_list *src, struct dm_list *dst) +{ + struct device_id_list *dil; + struct device_id_list *dil_new; + + dm_list_iterate_items(dil, src) { + if (!(dil_new = dm_pool_alloc(cmd->mem, sizeof(*dil_new)))) { + log_error("device_id_list alloc failed."); + return ECMD_FAILED; + } + + dil_new->dev = dil->dev; + strncpy(dil_new->pvid, dil->pvid, ID_LEN); + dm_list_add(dst, &dil_new->list); + } + + return ECMD_PROCESSED; +} + +/* + * For each device in arg_devices or all_devices that has a pvid, add a copy of + * that device to arg_missed. All PVs (devices with a pvid) should have been + * found while processing all VGs (including orphan VGs). But, some may have + * been missed if VGs were changing at the same time. This function creates a + * list of PVs that still remain in the given list, i.e. were missed the first + * time. A second iteration through VGs can look for these explicitly. + * (arg_devices is used if specific PVs are being processed; all_devices is + * used if all devs are being processed) + */ +static int _get_missed_pvs(struct cmd_context *cmd, + struct dm_list *devices, + struct dm_list *arg_missed) +{ + struct device_id_list *dil; + struct device_id_list *dil_missed; + + dm_list_iterate_items(dil, devices) { + if (!dil->pvid[0]) + continue; + + if (!(dil_missed = dm_pool_alloc(cmd->mem, sizeof(*dil_missed)))) { + log_error("device_id_list alloc failed."); + return ECMD_FAILED; + } + + dil_missed->dev = dil->dev; + strncpy(dil_missed->pvid, dil->pvid, ID_LEN); + dm_list_add(arg_missed, &dil_missed->list); + } + + return ECMD_PROCESSED; +} + static int _process_device_list(struct cmd_context *cmd, struct dm_list *all_devices, struct processing_handle *handle, process_single_pv_fn_t process_single_pv) @@ -2921,6 +2979,7 @@ static int _process_pvs_in_vgs(struct cmd_context *cmd, uint32_t read_flags, int ret_max = ECMD_PROCESSED; int ret; int skip; + int notfound; dm_list_iterate_items(vgnl, all_vgnameids) { if (sigint_caught()) @@ -2929,6 +2988,7 @@ static int _process_pvs_in_vgs(struct cmd_context *cmd, uint32_t read_flags, vg_name = vgnl->vg_name; vg_uuid = vgnl->vgid; skip = 0; + notfound = 0; if (!lockd_vg(cmd, vg_name, NULL, 0, &lockd_state)) { ret_max = ECMD_FAILED; @@ -2936,13 +2996,15 @@ static int _process_pvs_in_vgs(struct cmd_context *cmd, uint32_t read_flags, } vg = vg_read(cmd, vg_name, vg_uuid, read_flags, lockd_state); - if (_ignore_vg(vg, vg_name, NULL, read_flags, &skip)) { + if (_ignore_vg(vg, vg_name, NULL, read_flags, &skip, &notfound)) { stack; ret_max = ECMD_FAILED; if (!skip) goto endvg; /* Drop through to eliminate a clustered VG's PVs from the devices list */ } + if (notfound) + goto endvg; /* * Don't continue when skip is set, because we need to remove @@ -2982,6 +3044,7 @@ int process_each_pv(struct cmd_context *cmd, struct dm_list arg_tags; /* str_list */ struct dm_list arg_pvnames; /* str_list */ struct dm_list arg_devices; /* device_id_list */ + struct dm_list arg_missed; /* device_id_list */ struct dm_list all_vgnameids; /* vgnameid_list */ struct dm_list all_devices; /* device_id_list */ struct device_id_list *dil; @@ -3009,6 +3072,7 @@ int process_each_pv(struct cmd_context *cmd, dm_list_init(&arg_tags); dm_list_init(&arg_pvnames); dm_list_init(&arg_devices); + dm_list_init(&arg_missed); dm_list_init(&all_vgnameids); dm_list_init(&all_devices); @@ -3068,6 +3132,49 @@ int process_each_pv(struct cmd_context *cmd, if (ret > ret_max) ret_max = ret; + /* + * Some PVs may have been missed by the first search if another command + * moved them at the same time. Repeat the search for only the + * specific PVs missed. lvmcache needs clearing for a fresh search. + * + * If missed PVs are found in this repeated search, they are removed + * from the arg_missed list, but they also need to be removed from the + * arg_devices list, otherwise the check at the end will produce an + * error, thinking they weren't found. This is the reason for saving + * and comparing the original arg_missed list. + */ + if (!process_all_pvs) + _get_missed_pvs(cmd, &arg_devices, &arg_missed); + else + _get_missed_pvs(cmd, &all_devices, &arg_missed); + + if (!dm_list_empty(&arg_missed)) { + struct dm_list arg_missed_orig; + + dm_list_init(&arg_missed_orig); + _device_list_copy(cmd, &arg_missed, &arg_missed_orig); + + log_warn("Some PVs were not found in first search, retrying."); + + lvmcache_destroy(cmd, 0, 0); + lvmcache_init(); + lvmcache_seed_infos_from_lvmetad(cmd); + + ret = _process_pvs_in_vgs(cmd, read_flags, &all_vgnameids, &all_devices, + &arg_missed, &arg_tags, 0, 0, + handle, process_single_pv); + if (ret != ECMD_PROCESSED) + stack; + if (ret > ret_max) + ret_max = ret; + + /* Devices removed from arg_missed are removed from arg_devices. */ + dm_list_iterate_items(dil, &arg_missed_orig) { + if (!_device_list_find_dev(&arg_missed, dil->dev)) + _device_list_remove(&arg_devices, dil->dev); + } + } + dm_list_iterate_items(dil, &arg_devices) { log_error("Failed to find physical volume \"%s\".", dev_name(dil->dev)); ret_max = ECMD_FAILED;

1 0

master - tests: snapshot now respects threshold
by Zdenek Kabelac 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=b29593378f41f6dd… Commit: b29593378f41f6dd401fb9582fba1a100f33dbbb Parent: 428ca9b1207eba53ffc721bd9e528a1a148128c2 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 26 07:38:23 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Mon Oct 26 07:38:23 2015 +0100 tests: snapshot now respects threshold --- test/shell/snapshot-usage.sh | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/test/shell/snapshot-usage.sh b/test/shell/snapshot-usage.sh index ed9d8cb..d15eee1 100644 --- a/test/shell/snapshot-usage.sh +++ b/test/shell/snapshot-usage.sh @@ -169,13 +169,13 @@ lvcreate -s -l12 $vg1/lvol0 fill 1K check lv_field $vg1/lvol1 data_percent "100.00" -# Check it resizes 100% full valid snapshot +# Check it resizes 100% full valid snapshot to fit threshold lvextend --use-policies $vg1/lvol1 -check lv_field $vg1/lvol1 data_percent "80.00" +check lv_field $vg1/lvol1 data_percent "50.00" fill 4K lvextend --use-policies $vg1/lvol1 -check lv_field $vg1/lvol1 size "18.00k" +check lv_field $vg1/lvol1 size "24.00k" lvextend -l+33 $vg1/lvol1 check lv_field $vg1/lvol1 size "$EXPECT3"

1 0

master - libdm: enable no_flush for driver version > 11
by Zdenek Kabelac 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=428ca9b1207eba53… Commit: 428ca9b1207eba53ffc721bd9e528a1a148128c2 Parent: f898cf7539149a54a121b0aae59076c2d647bb37 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 26 07:37:59 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Mon Oct 26 07:37:59 2015 +0100 libdm: enable no_flush for driver version > 11 It appears the driver version 11 has troubles with usage of no_flush So require at least version 12. --- libdm/ioctl/libdm-iface.c | 9 +++++++-- 1 files changed, 7 insertions(+), 2 deletions(-) diff --git a/libdm/ioctl/libdm-iface.c b/libdm/ioctl/libdm-iface.c index 58a8dfe..2a36b77 100644 --- a/libdm/ioctl/libdm-iface.c +++ b/libdm/ioctl/libdm-iface.c @@ -1202,8 +1202,13 @@ static struct dm_ioctl *_flatten(struct dm_task *dmt, unsigned repeat_count) if (dmt->type == DM_DEVICE_SUSPEND) dmi->flags |= DM_SUSPEND_FLAG; - if (dmt->no_flush) - dmi->flags |= DM_NOFLUSH_FLAG; + if (dmt->no_flush) { + if (_dm_version_minor < 12) + log_verbose("No flush flag unsupported by kernel. " + "Buffers will be flushed."); + else + dmi->flags |= DM_NOFLUSH_FLAG; + } if (dmt->read_only) dmi->flags |= DM_READONLY_FLAG; if (dmt->skip_lockfs)

1 0

master - tests: snapshot now respects threshold
by Zdenek Kabelac 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=3466cd8a6caff3ca… Commit: 3466cd8a6caff3ca9ad00e22b927dbe22404fe5c Parent: 6e784abbd7336e57fee80dfb6a5a3e7f2fab7a98 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 26 07:34:33 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Mon Oct 26 07:34:33 2015 +0100 tests: snapshot now respects threshold --- test/shell/snapshot-usage.sh | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/test/shell/snapshot-usage.sh b/test/shell/snapshot-usage.sh index ed9d8cb..d15eee1 100644 --- a/test/shell/snapshot-usage.sh +++ b/test/shell/snapshot-usage.sh @@ -169,13 +169,13 @@ lvcreate -s -l12 $vg1/lvol0 fill 1K check lv_field $vg1/lvol1 data_percent "100.00" -# Check it resizes 100% full valid snapshot +# Check it resizes 100% full valid snapshot to fit threshold lvextend --use-policies $vg1/lvol1 -check lv_field $vg1/lvol1 data_percent "80.00" +check lv_field $vg1/lvol1 data_percent "50.00" fill 4K lvextend --use-policies $vg1/lvol1 -check lv_field $vg1/lvol1 size "18.00k" +check lv_field $vg1/lvol1 size "24.00k" lvextend -l+33 $vg1/lvol1 check lv_field $vg1/lvol1 size "$EXPECT3"

1 0

master - libdm: enable no_flush for driver version > 11
by Zdenek Kabelac 26 Oct '15

26 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=6e784abbd7336e57… Commit: 6e784abbd7336e57fee80dfb6a5a3e7f2fab7a98 Parent: f898cf7539149a54a121b0aae59076c2d647bb37 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 26 07:34:14 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Mon Oct 26 07:34:14 2015 +0100 libdm: enable no_flush for driver version > 11 It appears the driver version 11 has troubles with usage of no_flush So require at least version 12. --- libdm/ioctl/libdm-iface.c | 9 +++++++-- 1 files changed, 7 insertions(+), 2 deletions(-) diff --git a/libdm/ioctl/libdm-iface.c b/libdm/ioctl/libdm-iface.c index 58a8dfe..615bbd5 100644 --- a/libdm/ioctl/libdm-iface.c +++ b/libdm/ioctl/libdm-iface.c @@ -1202,8 +1202,13 @@ static struct dm_ioctl *_flatten(struct dm_task *dmt, unsigned repeat_count) if (dmt->type == DM_DEVICE_SUSPEND) dmi->flags |= DM_SUSPEND_FLAG; - if (dmt->no_flush) - dmi->flags |= DM_NOFLUSH_FLAG; + if (dmt->no_flush) { + if (_dm_version_minor < 11) + log_verbose("No flush flag unsupported by kernel. " + "Buffers will be flushed."); + else + dmi->flags |= DM_NOFLUSH_FLAG; + } if (dmt->read_only) dmi->flags |= DM_READONLY_FLAG; if (dmt->skip_lockfs)

1 0

master - tests: snapshot now respects threshold
by Zdenek Kabelac 25 Oct '15

25 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=76b49318b7b1d99f… Commit: 76b49318b7b1d99f16ecf8069d4cb1786bfc124f Parent: 9139d7a4bbaf0b6944802b7c6f893e23b59633f7 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sun Oct 25 23:44:58 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Sun Oct 25 23:44:58 2015 +0100 tests: snapshot now respects threshold --- test/shell/snapshot-usage.sh | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/test/shell/snapshot-usage.sh b/test/shell/snapshot-usage.sh index ed9d8cb..d15eee1 100644 --- a/test/shell/snapshot-usage.sh +++ b/test/shell/snapshot-usage.sh @@ -169,13 +169,13 @@ lvcreate -s -l12 $vg1/lvol0 fill 1K check lv_field $vg1/lvol1 data_percent "100.00" -# Check it resizes 100% full valid snapshot +# Check it resizes 100% full valid snapshot to fit threshold lvextend --use-policies $vg1/lvol1 -check lv_field $vg1/lvol1 data_percent "80.00" +check lv_field $vg1/lvol1 data_percent "50.00" fill 4K lvextend --use-policies $vg1/lvol1 -check lv_field $vg1/lvol1 size "18.00k" +check lv_field $vg1/lvol1 size "24.00k" lvextend -l+33 $vg1/lvol1 check lv_field $vg1/lvol1 size "$EXPECT3"

1 0

master - libdm: enable no_flush for driver version > 11
by Zdenek Kabelac 25 Oct '15

25 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=9139d7a4bbaf0b69… Commit: 9139d7a4bbaf0b6944802b7c6f893e23b59633f7 Parent: f898cf7539149a54a121b0aae59076c2d647bb37 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sun Oct 25 23:43:47 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Sun Oct 25 23:43:47 2015 +0100 libdm: enable no_flush for driver version > 11 It appears the driver version 11 has troubles with usage of no_flush So require at least version 12. --- libdm/ioctl/libdm-iface.c | 9 +++++++-- 1 files changed, 7 insertions(+), 2 deletions(-) diff --git a/libdm/ioctl/libdm-iface.c b/libdm/ioctl/libdm-iface.c index 58a8dfe..615bbd5 100644 --- a/libdm/ioctl/libdm-iface.c +++ b/libdm/ioctl/libdm-iface.c @@ -1202,8 +1202,13 @@ static struct dm_ioctl *_flatten(struct dm_task *dmt, unsigned repeat_count) if (dmt->type == DM_DEVICE_SUSPEND) dmi->flags |= DM_SUSPEND_FLAG; - if (dmt->no_flush) - dmi->flags |= DM_NOFLUSH_FLAG; + if (dmt->no_flush) { + if (_dm_version_minor < 11) + log_verbose("No flush flag unsupported by kernel. " + "Buffers will be flushed."); + else + dmi->flags |= DM_NOFLUSH_FLAG; + } if (dmt->read_only) dmi->flags |= DM_READONLY_FLAG; if (dmt->skip_lockfs)

1 0

master - dev_manager: no flush for extension
by Zdenek Kabelac 25 Oct '15

25 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=f898cf7539149a54… Commit: f898cf7539149a54a121b0aae59076c2d647bb37 Parent: 844b00958492224c448b30f89ae5532bb6591fd2 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sun Oct 25 20:41:19 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Sun Oct 25 21:09:31 2015 +0100 dev_manager: no flush for extension Recognize the target only 'extends' and do not enforce 'flush' in this case. Only the size reduction still requires flush (so disables usage of no_flush flag). If some other targets do require flush before suspend, they have to explicitly ask for it. --- WHATS_NEW | 1 + lib/activate/dev_manager.c | 2 +- 2 files changed, 2 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 33d11a4..80a198a 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Enforce flush for suspend only when volume size is reduced. Enable code which detects the need of flush during suspend. Ensure --use-policy will resize volume to fit below threshold. Correct percentage evaluation when checking thin-pool over threshold. diff --git a/lib/activate/dev_manager.c b/lib/activate/dev_manager.c index 1caa2d7..dc6d2fd 100644 --- a/lib/activate/dev_manager.c +++ b/lib/activate/dev_manager.c @@ -3296,7 +3296,7 @@ static int _tree_action(struct dev_manager *dm, const struct logical_volume *lv, if (!dm_tree_preload_children(root, dlid, DLID_SIZE)) goto_out; - if (dm_tree_node_size_changed(root)) + if ((dm_tree_node_size_changed(root) < 0)) dm->flush_required = 1; if (action == ACTIVATE) {

1 0

master - dev_manager: enabled no_flush for suspend
by Zdenek Kabelac 25 Oct '15

25 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=844b00958492224c… Commit: 844b00958492224c448b30f89ae5532bb6591fd2 Parent: 9ef820a2a5e174cd08097981d7879e1f4cbded01 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sun Oct 25 20:37:39 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Sun Oct 25 21:07:37 2015 +0100 dev_manager: enabled no_flush for suspend While the activation code tries to evaluate which target really needs flush with suspend and which may go without flush, it has stayed effectively disabled by original commit: 33f732c5e9493cda4b161a18b3d53885d207e3b8 since here it only allows to pass non-pvmoving 'mirrors'. So remove check for mirror LV type and only disable no_flush for 'pvmove'.. TODO: Looking into history - it also seemed like raid target would have always required flushing but it's been later removed without clean explanation. If some more targets really do need 'no_flush' it should been handle at their 'level' - since we now stack multiple targets over itself. --- WHATS_NEW | 1 + lib/activate/dev_manager.c | 2 +- 2 files changed, 2 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 8ca0aa7..33d11a4 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Enable code which detects the need of flush during suspend. Ensure --use-policy will resize volume to fit below threshold. Correct percentage evaluation when checking thin-pool over threshold. Fix lvmcache to move PV from VG to orphans if VG is removed and lvmetad used. diff --git a/lib/activate/dev_manager.c b/lib/activate/dev_manager.c index a5ee824..1caa2d7 100644 --- a/lib/activate/dev_manager.c +++ b/lib/activate/dev_manager.c @@ -3277,7 +3277,7 @@ static int _tree_action(struct dev_manager *dm, const struct logical_volume *lv, break; case SUSPEND: dm_tree_skip_lockfs(root); - if (!dm->flush_required && lv_is_mirror(lv) && !lv_is_pvmove(lv)) + if (!dm->flush_required && !lv_is_pvmove(lv)) dm_tree_use_no_flush_suspend(root); /* Fall through */ case SUSPEND_WITH_LOCKFS:

1 0

master - libdm: dm_tree_node_size_changed recognizes reduction
by Zdenek Kabelac 25 Oct '15

25 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=9ef820a2a5e174cd… Commit: 9ef820a2a5e174cd08097981d7879e1f4cbded01 Parent: 40eea582aea512db7fceafd8559bf66df0733b55 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sun Oct 25 19:27:09 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Sun Oct 25 21:05:15 2015 +0100 libdm: dm_tree_node_size_changed recognizes reduction Add more functionality to size_changed function. While 'existing' API only detected 0 for unchanged, and !0 for changed, new improved API will also detected if the size has only went bigger - or there was size reduction. Function work for the whole dm-tree - so no change is size is always 0. only size extension 1. and if some size reduction is there - returns -1. This result can be used for better evaluation whether we need to flush before suspend. --- WHATS_NEW_DM | 1 + libdm/.exported_symbols.Base | 1 - libdm/.exported_symbols.DM_1_02_110 | 1 + libdm/libdevmapper.h | 5 +++++ libdm/libdm-deptree.c | 25 ++++++++++++++++++++++--- 5 files changed, 29 insertions(+), 4 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 910f6d6..e09ab63 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Enhance dm_tree_node_size_changed() to recognize size reduction. Support exit on idle for dmenventd (1 hour). Add support to allow unmonitor device from plugin itself. New design for thread co-operation in dmeventd. diff --git a/libdm/.exported_symbols.Base b/libdm/.exported_symbols.Base index f9c3cb1..27fef53 100644 --- a/libdm/.exported_symbols.Base +++ b/libdm/.exported_symbols.Base @@ -262,7 +262,6 @@ dm_tree_node_set_thin_external_origin dm_tree_node_set_thin_pool_discard dm_tree_node_set_thin_pool_error_if_no_space dm_tree_node_set_udev_flags -dm_tree_node_size_changed dm_tree_preload_children dm_tree_retry_remove dm_tree_set_cookie diff --git a/libdm/.exported_symbols.DM_1_02_110 b/libdm/.exported_symbols.DM_1_02_110 index da742ee..eba5625 100644 --- a/libdm/.exported_symbols.DM_1_02_110 +++ b/libdm/.exported_symbols.DM_1_02_110 @@ -1,2 +1,3 @@ dm_report_compact_given_fields dm_hold_control_dev +dm_tree_node_size_changed diff --git a/libdm/libdevmapper.h b/libdm/libdevmapper.h index 08b24fe..f7ab257 100644 --- a/libdm/libdevmapper.h +++ b/libdm/libdevmapper.h @@ -1256,6 +1256,11 @@ const char *dm_tree_node_get_name(const struct dm_tree_node *node); const char *dm_tree_node_get_uuid(const struct dm_tree_node *node); const struct dm_info *dm_tree_node_get_info(const struct dm_tree_node *node); void *dm_tree_node_get_context(const struct dm_tree_node *node); +/* + * Returns 0 when node size and its children is unchanged. + * Returns 1 when node or any of its children has increased size. + * Rerurns -1 when node or any of its children has reduced size. + */ int dm_tree_node_size_changed(const struct dm_tree_node *dnode); /* diff --git a/libdm/libdm-deptree.c b/libdm/libdm-deptree.c index afdce8d..d7e2143 100644 --- a/libdm/libdm-deptree.c +++ b/libdm/libdm-deptree.c @@ -220,7 +220,7 @@ struct load_properties { uint32_t read_ahead_flags; unsigned segment_count; - unsigned size_changed; + int size_changed; struct dm_list segs; const char *new_name; @@ -2729,7 +2729,8 @@ static int _load_node(struct dm_tree_node *dnode) existing_table_size = dm_task_get_existing_table_size(dmt); if ((dnode->props.size_changed = - (existing_table_size == seg_start) ? 0 : 1)) { + (existing_table_size == seg_start) ? 0 : + (existing_table_size > seg_start) ? -1 : 1)) { /* * Kernel usually skips size validation on zero-length devices * now so no need to preload them. @@ -2825,8 +2826,10 @@ int dm_tree_preload_children(struct dm_tree_node *dnode, } /* Propagate device size change change */ - if (child->props.size_changed) + if (child->props.size_changed > 0 && !dnode->props.size_changed) dnode->props.size_changed = 1; + else if (child->props.size_changed < 0) + dnode->props.size_changed = -1; /* Resume device immediately if it has parents and its size changed */ if (!dm_tree_node_num_children(child, 1) || !child->props.size_changed) @@ -3853,3 +3856,19 @@ void dm_tree_node_set_callback(struct dm_tree_node *dnode, dnode->callback = cb; dnode->callback_data = data; } + +/* + * Backward compatible dm_tree_node_size_changed() implementations. + * + * Keep these at the end of the file to avoid adding clutter around the + * current dm_tree_node_size_changed() version. + */ +#if defined(__GNUC__) +int dm_tree_node_size_changed_base(const struct dm_tree_node *dnode); +DM_EXPORT_SYMBOL_BASE(dm_tree_node_size_changed); +int dm_tree_node_size_changed_base(const struct dm_tree_node *dnode) +{ + /* Base does not make difference between smaller and bigger */ + return dm_tree_node_size_changed(dnode) ? 1 : 0; +} +#endif

1 0

master - lv_manip: ensure it will fit bellow threshold
by Zdenek Kabelac 25 Oct '15

25 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=40eea582aea512db… Commit: 40eea582aea512db7fceafd8559bf66df0733b55 Parent: b780d329aa9ce249bf5164154c81cefd9b276970 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sun Oct 25 19:19:39 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Sun Oct 25 21:03:11 2015 +0100 lv_manip: ensure it will fit bellow threshold Use single code to evaluate if the percentage value has crossed threshold. Recalculate amount value to always fit bellow threshold so there are not need any extra reiterations to reach this state in case policy amount is too small. --- WHATS_NEW | 1 + lib/metadata/lv_manip.c | 37 +++++++++++++++++++++++++------------ 2 files changed, 26 insertions(+), 12 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 0495e46..8ca0aa7 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Ensure --use-policy will resize volume to fit below threshold. Correct percentage evaluation when checking thin-pool over threshold. Fix lvmcache to move PV from VG to orphans if VG is removed and lvmetad used. Fix lvmcache to not cache even invalid info about PV which got removed. diff --git a/lib/metadata/lv_manip.c b/lib/metadata/lv_manip.c index 19032e9..8029dab 100644 --- a/lib/metadata/lv_manip.c +++ b/lib/metadata/lv_manip.c @@ -4424,6 +4424,25 @@ static int _fsadm_cmd(struct cmd_context *cmd, return exec_cmd(cmd, argv, status, 1); } +static int _adjust_amount(dm_percent_t percent, int policy_threshold, int *policy_amount) +{ + if (!(DM_PERCENT_0 < percent && percent <= DM_PERCENT_100) || + percent <= (policy_threshold * DM_PERCENT_1)) + return 0; + /* + * Evaluate the minimal amount needed to get bellow threshold. + * Keep using DM_PERCENT_1 units for better precision. + * Round-up to needed percentage value + */ + percent = (percent/policy_threshold + (DM_PERCENT_1 - 1) / 100) / (DM_PERCENT_1 / 100) - 100; + + /* Use it if current policy amount is smaller */ + if (*policy_amount < percent) + *policy_amount = percent; + + return 1; +} + static int _adjust_policy_params(struct cmd_context *cmd, struct logical_volume *lv, struct lvresize_params *lp) { @@ -4469,35 +4488,29 @@ static int _adjust_policy_params(struct cmd_context *cmd, return 0; } - policy_threshold *= DM_PERCENT_1; - if (lv_is_thin_pool(lv)) { if (!lv_thin_pool_percent(lv, 1, &percent)) return_0; - if ((DM_PERCENT_0 < percent && percent <= DM_PERCENT_100) && - (percent > policy_threshold)) { + if (_adjust_amount(percent, policy_threshold, &policy_amount)) { if (!thin_pool_feature_supported(lv, THIN_FEATURE_METADATA_RESIZE)) { - log_error_once("Online metadata resize for %s/%s is not supported.", - lv->vg->name, lv->name); + log_error_once("Online metadata resize for %s is not supported.", + display_lvname(lv)); return 0; } lp->poolmetadatasize = (first_seg(lv)->metadata_lv->size * policy_amount + 99) / 100; lp->poolmetadatasign = SIGN_PLUS; } - if (!lv_thin_pool_percent(lv, 0, &percent)) return_0; - if (!(DM_PERCENT_0 < percent && percent <= DM_PERCENT_100) || - percent <= policy_threshold) - return 1; } else { if (!lv_snapshot_percent(lv, &percent)) return_0; - if (!(DM_PERCENT_0 < percent && percent <= DM_PERCENT_100) || percent <= policy_threshold) - return 1; /* nothing to do */ } + if (!_adjust_amount(percent, policy_threshold, &policy_amount)) + return 1; /* nothing to do */ + lp->extents = policy_amount; lp->sizeargs = (lp->extents) ? 1 : 0;

1 0

master - thin: fix percentage compare
by Zdenek Kabelac 25 Oct '15

25 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=b780d329aa9ce249… Commit: b780d329aa9ce249bf5164154c81cefd9b276970 Parent: 8f269697d21035035ffb4954f31bd702a6dd7ca8 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sun Oct 25 19:13:53 2015 +0100 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Sun Oct 25 21:01:54 2015 +0100 thin: fix percentage compare Since plugin's percentage compare has been fixed, it's now revealed wrong compare here. The logic for threshold is - to allow to go as high as given value e.g. 80% - so if pool is exactlu 80% full it's still allowed to use it (dmeventd will not resize it). --- WHATS_NEW | 1 + lib/metadata/thin_manip.c | 4 ++-- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 5911c64..0495e46 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Correct percentage evaluation when checking thin-pool over threshold. Fix lvmcache to move PV from VG to orphans if VG is removed and lvmetad used. Fix lvmcache to not cache even invalid info about PV which got removed. Support checking of memlock daemon counter. diff --git a/lib/metadata/thin_manip.c b/lib/metadata/thin_manip.c index 37bf4cb..6b22ce5 100644 --- a/lib/metadata/thin_manip.c +++ b/lib/metadata/thin_manip.c @@ -224,7 +224,7 @@ int pool_below_threshold(const struct lv_segment *pool_seg) if (!lv_thin_pool_percent(pool_seg->lv, 0, &percent)) return_0; - if (percent >= threshold) { + if (percent > threshold) { log_debug("Threshold configured for free data space in " "thin pool %s has been reached (%.2f%% >= %.2f%%).", display_lvname(pool_seg->lv), @@ -237,7 +237,7 @@ int pool_below_threshold(const struct lv_segment *pool_seg) if (!lv_thin_pool_percent(pool_seg->lv, 1, &percent)) return_0; - if (percent >= threshold) { + if (percent > threshold) { log_debug("Threshold configured for free metadata space in " "thin pool %s has been reached (%.2f%% > %.2f%%).", display_lvname(pool_seg->lv),

1 0

master - vg_read: remove unused inconsistent check
by David Teigland 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=8f269697d2103503… Commit: 8f269697d21035035ffb4954f31bd702a6dd7ca8 Parent: 98d81a43ea31ec963a36b8610f3b5a9be28b4cc1 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Fri Oct 23 12:19:11 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Fri Oct 23 12:19:11 2015 -0500 vg_read: remove unused inconsistent check Commit 1a74171ca5682a684d0e05c6090c3d33cab8795b added a check to ignore a VG that was FAILED_INCONSISTENT if the command doesn't care if the VG is not found. Remove that check because that case is never reached by the current code. --- tools/toollib.c | 6 ------ 1 files changed, 0 insertions(+), 6 deletions(-) diff --git a/tools/toollib.c b/tools/toollib.c index 39bf0ac..fba075a 100644 --- a/tools/toollib.c +++ b/tools/toollib.c @@ -195,12 +195,6 @@ static int _ignore_vg(struct volume_group *vg, const char *vg_name, return 0; } - if ((read_error & FAILED_INCONSISTENT) && (read_flags & READ_OK_NOTFOUND)) { - read_error &= ~FAILED_INCONSISTENT; - *skip = 1; - return 0; - } - if ((read_error & FAILED_INCONSISTENT) && (read_flags & READ_ALLOW_INCONSISTENT)) read_error &= ~FAILED_INCONSISTENT; /* Check for other errors */

1 0

master - vgextend: fix use of the wrong flag
by David Teigland 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=98d81a43ea31ec96… Commit: 98d81a43ea31ec963a36b8610f3b5a9be28b4cc1 Parent: 1a74171ca5682a684d0e05c6090c3d33cab8795b Author: David Teigland <teigland(a)redhat.com> AuthorDate: Fri Oct 23 10:21:06 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Fri Oct 23 10:21:06 2015 -0500 vgextend: fix use of the wrong flag The ONE_VGNAME_ARG was being passed and tested as vg_read() flag but it's a cmd struct flag. (It affects command arg processing in toollib, not vg_read behavior. Flags related to command processing are generally cmd struct flags, while vg_read arg flags are generally related to vg_read behavior.) --- tools/commands.h | 2 +- tools/toollib.c | 2 +- tools/vgextend.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/tools/commands.h b/tools/commands.h index 92d13fd..ae5dff7 100644 --- a/tools/commands.h +++ b/tools/commands.h @@ -1211,7 +1211,7 @@ xx(vgexport, xx(vgextend, "Add physical volumes to a volume group", - 0, + ONE_VGNAME_ARG, "vgextend\n" "\t[-A|--autobackup y|n]\n" "\t[--restoremissing]\n" diff --git a/tools/toollib.c b/tools/toollib.c index b17fe53..39bf0ac 100644 --- a/tools/toollib.c +++ b/tools/toollib.c @@ -2032,7 +2032,7 @@ int process_each_vg(struct cmd_context *cmd, int argc, char **argv, struct dm_list vgnameids_to_process; /* vgnameid_list */ int enable_all_vgs = (cmd->command->flags & ALL_VGS_IS_DEFAULT); - unsigned one_vgname_arg = (flags & ONE_VGNAME_ARG); + int one_vgname_arg = (cmd->command->flags & ONE_VGNAME_ARG); int ret; /* Disable error in vg_read so we can print it from ignore_vg. */ diff --git a/tools/vgextend.c b/tools/vgextend.c index 581c211..bddc22f 100644 --- a/tools/vgextend.c +++ b/tools/vgextend.c @@ -170,7 +170,7 @@ int vgextend(struct cmd_context *cmd, int argc, char **argv) return_ECMD_FAILED; ret = process_each_vg(cmd, argc, argv, - READ_FOR_UPDATE | ONE_VGNAME_ARG, handle, + READ_FOR_UPDATE, handle, restoremissing ? &_vgextend_restoremissing : &_vgextend_single); destroy_processing_handle(cmd, handle);

1 0

master - vg_read: sometimes ignore read errors
by David Teigland 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=1a74171ca5682a68… Commit: 1a74171ca5682a684d0e05c6090c3d33cab8795b Parent: 51735f09f79794d6609913b4c117f1a4a0e491fe Author: David Teigland <teigland(a)redhat.com> AuthorDate: Thu Oct 22 14:56:22 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Fri Oct 23 10:12:34 2015 -0500 vg_read: sometimes ignore read errors Running "vgremove -f VG & pvs" results in the pvs command reporting that the VG is not found or is inconsistent. If the VG is gone or being removed, the pvs command should just skip it and not print errors about it. "Not found" is because the pvs command created the list of VGs to process, including VG, then vgremove removed the VG, then the pvs command came to to read the VG to process it and did not find it. An "inconsistent" error could be reported if vgremove had only partially completed removing VG when pvs did vg_read on the VG to process it, causing pvs to find the VG in a partially-removed state. This fix adds a flag that pvs uses to ignore a VG that can't be read or is inconsistent. --- lib/metadata/metadata-exported.h | 9 +++---- lib/metadata/metadata.c | 33 +++++++++++++++------------- tools/toollib.c | 43 ++++++++++++++++++++++++++++++-------- 3 files changed, 56 insertions(+), 29 deletions(-) diff --git a/lib/metadata/metadata-exported.h b/lib/metadata/metadata-exported.h index 25320fd..41838a8 100644 --- a/lib/metadata/metadata-exported.h +++ b/lib/metadata/metadata-exported.h @@ -163,10 +163,9 @@ /* vg_read and vg_read_for_update flags */ #define READ_ALLOW_INCONSISTENT 0x00010000U #define READ_ALLOW_EXPORTED 0x00020000U +#define READ_OK_NOTFOUND 0x00040000U #define READ_WARN_INCONSISTENT 0x00080000U - -/* A meta-flag, useful with toollib for_each_* functions. */ -#define READ_FOR_UPDATE 0x00100000U +#define READ_FOR_UPDATE 0x00100000U /* A meta-flag, useful with toollib for_each_* functions. */ /* vg's "read_status" field */ #define FAILED_INCONSISTENT 0x00000001U @@ -650,9 +649,9 @@ int lv_resize(struct cmd_context *cmd, struct logical_volume *lv, * Return a handle to VG metadata. */ struct volume_group *vg_read(struct cmd_context *cmd, const char *vg_name, - const char *vgid, uint32_t flags, uint32_t lockd_state); + const char *vgid, uint32_t read_flags, uint32_t lockd_state); struct volume_group *vg_read_for_update(struct cmd_context *cmd, const char *vg_name, - const char *vgid, uint32_t flags, uint32_t lockd_state); + const char *vgid, uint32_t read_flags, uint32_t lockd_state); /* * Test validity of a VG handle. diff --git a/lib/metadata/metadata.c b/lib/metadata/metadata.c index 9bd42c9..1db8f58 100644 --- a/lib/metadata/metadata.c +++ b/lib/metadata/metadata.c @@ -4925,7 +4925,7 @@ static int _vg_access_permitted(struct cmd_context *cmd, struct volume_group *vg * Consolidated locking, reading, and status flag checking. * * If the metadata is inconsistent, setting READ_ALLOW_INCONSISTENT in - * misc_flags will return it with FAILED_INCONSISTENT set instead of + * read_flags will return it with FAILED_INCONSISTENT set instead of * giving you nothing. * * Use vg_read_error(vg) to determine the result. Nonzero means there were @@ -4933,8 +4933,10 @@ static int _vg_access_permitted(struct cmd_context *cmd, struct volume_group *vg * Zero value means that the VG is open and appropriate locks are held. */ static struct volume_group *_vg_lock_and_read(struct cmd_context *cmd, const char *vg_name, - const char *vgid, uint32_t lock_flags, - uint64_t status_flags, uint32_t misc_flags, + const char *vgid, + uint32_t lock_flags, + uint64_t status_flags, + uint32_t read_flags, uint32_t lockd_state) { struct volume_group *vg = NULL; @@ -4944,7 +4946,7 @@ static struct volume_group *_vg_lock_and_read(struct cmd_context *cmd, const cha uint32_t warn_flags = 0; int already_locked; - if (misc_flags & READ_ALLOW_INCONSISTENT || lock_flags != LCK_VG_WRITE) + if ((read_flags & READ_ALLOW_INCONSISTENT) || (lock_flags != LCK_VG_WRITE)) consistent = 0; if (!validate_name(vg_name) && !is_orphan_vg(vg_name)) { @@ -4967,7 +4969,7 @@ static struct volume_group *_vg_lock_and_read(struct cmd_context *cmd, const cha consistent_in = consistent; warn_flags = WARN_PV_READ; - if (consistent || (misc_flags & READ_WARN_INCONSISTENT)) + if (consistent || (read_flags & READ_WARN_INCONSISTENT)) warn_flags |= WARN_INCONSISTENT; /* If consistent == 1, we get NULL here if correction fails. */ @@ -4976,7 +4978,8 @@ static struct volume_group *_vg_lock_and_read(struct cmd_context *cmd, const cha failure |= FAILED_INCONSISTENT; goto bad; } - log_error("Volume group \"%s\" not found", vg_name); + if (!(read_flags & READ_OK_NOTFOUND)) + log_error("Volume group \"%s\" not found", vg_name); failure |= FAILED_NOTFOUND; goto bad; } @@ -5057,20 +5060,20 @@ bad_no_unlock: * *consistent = 1. */ struct volume_group *vg_read(struct cmd_context *cmd, const char *vg_name, - const char *vgid, uint32_t flags, uint32_t lockd_state) + const char *vgid, uint32_t read_flags, uint32_t lockd_state) { - uint64_t status = UINT64_C(0); + uint64_t status_flags = UINT64_C(0); uint32_t lock_flags = LCK_VG_READ; - if (flags & READ_FOR_UPDATE) { - status |= EXPORTED_VG | LVM_WRITE; + if (read_flags & READ_FOR_UPDATE) { + status_flags |= EXPORTED_VG | LVM_WRITE; lock_flags = LCK_VG_WRITE; } - if (flags & READ_ALLOW_EXPORTED) - status &= ~EXPORTED_VG; + if (read_flags & READ_ALLOW_EXPORTED) + status_flags &= ~EXPORTED_VG; - return _vg_lock_and_read(cmd, vg_name, vgid, lock_flags, status, flags, lockd_state); + return _vg_lock_and_read(cmd, vg_name, vgid, lock_flags, status_flags, read_flags, lockd_state); } /* @@ -5079,9 +5082,9 @@ struct volume_group *vg_read(struct cmd_context *cmd, const char *vg_name, * request the new metadata to be written and committed). */ struct volume_group *vg_read_for_update(struct cmd_context *cmd, const char *vg_name, - const char *vgid, uint32_t flags, uint32_t lockd_state) + const char *vgid, uint32_t read_flags, uint32_t lockd_state) { - return vg_read(cmd, vg_name, vgid, flags | READ_FOR_UPDATE, lockd_state); + return vg_read(cmd, vg_name, vgid, read_flags | READ_FOR_UPDATE, lockd_state); } /* diff --git a/tools/toollib.c b/tools/toollib.c index dfb2b87..b17fe53 100644 --- a/tools/toollib.c +++ b/tools/toollib.c @@ -184,12 +184,24 @@ const char *skip_dev_dir(struct cmd_context *cmd, const char *vg_name, * If *skip is 1, it's OK for the caller to read the list of PVs in the VG. */ static int _ignore_vg(struct volume_group *vg, const char *vg_name, - struct dm_list *arg_vgnames, int allow_inconsistent, int *skip) + struct dm_list *arg_vgnames, uint32_t read_flags, int *skip) { uint32_t read_error = vg_read_error(vg); *skip = 0; - if ((read_error & FAILED_INCONSISTENT) && allow_inconsistent) + if ((read_error & FAILED_NOTFOUND) && (read_flags & READ_OK_NOTFOUND)) { + read_error &= ~FAILED_NOTFOUND; + *skip = 1; + return 0; + } + + if ((read_error & FAILED_INCONSISTENT) && (read_flags & READ_OK_NOTFOUND)) { + read_error &= ~FAILED_INCONSISTENT; + *skip = 1; + return 0; + } + + if ((read_error & FAILED_INCONSISTENT) && (read_flags & READ_ALLOW_INCONSISTENT)) read_error &= ~FAILED_INCONSISTENT; /* Check for other errors */ if ((read_error & FAILED_CLUSTERED) && vg->cmd->ignore_clustered_vgs) { @@ -1943,7 +1955,7 @@ static int _process_vgnameid_list(struct cmd_context *cmd, uint32_t flags, } vg = vg_read(cmd, vg_name, vg_uuid, flags, lockd_state); - if (_ignore_vg(vg, vg_name, arg_vgnames, flags & READ_ALLOW_INCONSISTENT, &skip)) { + if (_ignore_vg(vg, vg_name, arg_vgnames, flags, &skip)) { stack; ret_max = ECMD_FAILED; goto endvg; @@ -2431,7 +2443,7 @@ static int _process_lv_vgnameid_list(struct cmd_context *cmd, uint32_t flags, } vg = vg_read(cmd, vg_name, vg_uuid, flags, lockd_state); - if (_ignore_vg(vg, vg_name, arg_vgnames, flags & READ_ALLOW_INCONSISTENT, &skip)) { + if (_ignore_vg(vg, vg_name, arg_vgnames, flags, &skip)) { stack; ret_max = ECMD_FAILED; goto endvg; @@ -2897,7 +2909,7 @@ out: * should produce an error. Any devices remaining in all_devices were * not found and should be processed by process_device_list(). */ -static int _process_pvs_in_vgs(struct cmd_context *cmd, uint32_t flags, +static int _process_pvs_in_vgs(struct cmd_context *cmd, uint32_t read_flags, struct dm_list *all_vgnameids, struct dm_list *all_devices, struct dm_list *arg_devices, @@ -2929,8 +2941,8 @@ static int _process_pvs_in_vgs(struct cmd_context *cmd, uint32_t flags, continue; } - vg = vg_read(cmd, vg_name, vg_uuid, flags | READ_WARN_INCONSISTENT, lockd_state); - if (_ignore_vg(vg, vg_name, NULL, flags & READ_ALLOW_INCONSISTENT, &skip)) { + vg = vg_read(cmd, vg_name, vg_uuid, read_flags, lockd_state); + if (_ignore_vg(vg, vg_name, NULL, read_flags, &skip)) { stack; ret_max = ECMD_FAILED; if (!skip) @@ -2969,7 +2981,7 @@ endvg: int process_each_pv(struct cmd_context *cmd, int argc, char **argv, const char *only_this_vgname, - uint32_t flags, + uint32_t read_flags, struct processing_handle *handle, process_single_pv_fn_t process_single_pv) { @@ -2984,6 +2996,19 @@ int process_each_pv(struct cmd_context *cmd, int ret_max = ECMD_PROCESSED; int ret; + /* + * When processing a specific VG name, warn if it's inconsistent and + * print an error if it's not found. Otherwise we're processing all + * VGs, in which case the command doesn't care if the VG is inconsisent + * or not found; it just wants to skip that VG. (It may be not found + * if it was removed between creating the list of all VGs and then + * processing each VG. + */ + if (only_this_vgname) + read_flags |= READ_WARN_INCONSISTENT; + else + read_flags |= READ_OK_NOTFOUND; + /* Disable error in vg_read so we can print it from ignore_vg. */ cmd->vg_read_print_access_error = 0; @@ -3040,7 +3065,7 @@ int process_each_pv(struct cmd_context *cmd, /* get_arg_devices reports the error for any PV names not found. */ ret_max = ECMD_FAILED; - ret = _process_pvs_in_vgs(cmd, flags, &all_vgnameids, &all_devices, + ret = _process_pvs_in_vgs(cmd, read_flags, &all_vgnameids, &all_devices, &arg_devices, &arg_tags, process_all_pvs, process_all_devices, handle, process_single_pv);

1 0

master - thin: Fix typo in policy threshold message.
by Alasdair Kergon 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=51735f09f79794d6… Commit: 51735f09f79794d6609913b4c117f1a4a0e491fe Parent: 3a42c13ccf151c3afdeab362c8541b5d249fb01e Author: Alasdair G Kergon <agk(a)redhat.com> AuthorDate: Fri Oct 23 15:38:31 2015 +0100 Committer: Alasdair G Kergon <agk(a)redhat.com> CommitterDate: Fri Oct 23 15:38:31 2015 +0100 thin: Fix typo in policy threshold message. --- lib/metadata/lv_manip.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/lib/metadata/lv_manip.c b/lib/metadata/lv_manip.c index 738343f..19032e9 100644 --- a/lib/metadata/lv_manip.c +++ b/lib/metadata/lv_manip.c @@ -4438,8 +4438,8 @@ static int _adjust_policy_params(struct cmd_context *cmd, find_config_tree_int(cmd, activation_thin_pool_autoextend_percent_CFG, lv_config_profile(lv)); if (policy_threshold < 50) { - log_warn("WARNING: Thin pool autoextend threshold %d%% is set bellow " - "minimal supported value 50%%.", policy_threshold); + log_warn("WARNING: Thin pool autoextend threshold %d%% is set below " + "minimum supported 50%%.", policy_threshold); policy_threshold = 50; } } else {

1 0

master - lvmcache: update cached info properly when moving from VG to orphan while lvmetad is used
by Peter Rajnoha 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=3a42c13ccf151c3a… Commit: 3a42c13ccf151c3afdeab362c8541b5d249fb01e Parent: 8b965bd3d5df808aef2012494602cc0cf73d7398 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 23 15:05:48 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 23 15:58:31 2015 +0200 lvmcache: update cached info properly when moving from VG to orphan while lvmetad is used When lvmetad is used and lvmcache update function (lvmcache_update_vgname_and_id) was called to update existing lvmcache records, a condition was met which made to retun from the update function immediately, effectively making it NOOP. It seems there's no reason for such condition and lvmcache should be update appropriately even when lvmetad used as lvmcache may be reused, most notably in lvm shell. It's possible this is a remnant of the lvmetad development code which didn't get removed for some reason and the bug didn't get spotted because lvm shell is not used often (the condition dates back to 2012 or so). Example, lvmetad and lvm shell used: lvm> pvs PV VG Fmt Attr PSize PFree /dev/sda vg lvm2 a-- 124.00m 124.00m Before this patch: ================== lvm> vgremove vg Volume group "vg" successfully removed lvm> pvs With this patch applied: ======================== lvm> vgremove vg Volume group "vg" successfully removed lvm> pvs PV VG Fmt Attr PSize PFree /dev/sda lvm2 --- 128.00m 128.00m --- WHATS_NEW | 1 + lib/cache/lvmcache.c | 4 ---- 2 files changed, 1 insertions(+), 4 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 6462e18..5911c64 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Fix lvmcache to move PV from VG to orphans if VG is removed and lvmetad used. Fix lvmcache to not cache even invalid info about PV which got removed. Support checking of memlock daemon counter. Allow all log levels to be used with the lvmetad -l option. diff --git a/lib/cache/lvmcache.c b/lib/cache/lvmcache.c index 1bc4241..31f2aa9 100644 --- a/lib/cache/lvmcache.c +++ b/lib/cache/lvmcache.c @@ -1542,10 +1542,6 @@ int lvmcache_update_vgname_and_id(struct lvmcache_info *info, struct lvmcache_vg vgid = vgname; } - /* When using lvmetad, the PV could not have become orphaned. */ - if (lvmetad_active() && is_orphan_vg(vgname) && info->vginfo) - return 1; - /* If PV without mdas is already in a real VG, don't make it orphan */ if (is_orphan_vg(vgname) && info->vginfo && mdas_empty_or_ignored(&info->mdas) &&

1 0

master - pvremove: make sure even invalid info is removed from lvmcache on pvremove
by Peter Rajnoha 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=8b965bd3d5df808a… Commit: 8b965bd3d5df808aef2012494602cc0cf73d7398 Parent: 1a7bea0f0f187a3d3332dda2cca2ecd0300b8031 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 23 14:11:16 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 23 15:48:06 2015 +0200 pvremove: make sure even invalid info is removed from lvmcache on pvremove The lvmcache info might be resued, most notably in lvm shell. We need to be sure that even lvmcache_info marked as invalid is removed from the lvmcache so it does not confuse any subsequent code/commands executed later on. Problematic example with the lvm shell: lvm> pvs PV VG Fmt Attr PSize PFree /dev/sda lvm2 --- 128.00m 128.00m Before this patch (/dev/sda still displayed in a way): ====================================================== lvm> pvremove /dev/sda Labels on physical volume "/dev/sda" successfully wiped (without lvmetad) lvm> pvs No physical volume label read from /dev/sda (with lvmetad) lvm> pvs PV VG Fmt Attr PSize PFree /dev/sda lvm2 --- 128.00m 128.00m With this patch applied: ======================== lvm> pvremove /dev/sda Labels on physical volume "/dev/sda" successfully wiped (without lvmetad) lvm> pvs (with lvmetad) lvm> pvs --- WHATS_NEW | 1 + lib/metadata/pv_manip.c | 2 +- 2 files changed, 2 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 8ffbeca..6462e18 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Fix lvmcache to not cache even invalid info about PV which got removed. Support checking of memlock daemon counter. Allow all log levels to be used with the lvmetad -l option. Add optional shutdown when idle support for lvmetad. diff --git a/lib/metadata/pv_manip.c b/lib/metadata/pv_manip.c index ce7f661..e48fe42 100644 --- a/lib/metadata/pv_manip.c +++ b/lib/metadata/pv_manip.c @@ -781,7 +781,7 @@ int pvremove_single(struct cmd_context *cmd, const char *pv_name, goto out; } - info = lvmcache_info_from_pvid(dev->pvid, 1); + info = lvmcache_info_from_pvid(dev->pvid, 0); if (!dev_test_excl(dev)) { /* FIXME Detect whether device-mapper is still using the device */

1 0

master - cleanup: drop debug
by Zdenek Kabelac 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=1a7bea0f0f187a3d… Commit: 1a7bea0f0f187a3d3332dda2cca2ecd0300b8031 Parent: b5b2a5483450dff10fd23a7b712649729e619b0d Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 23 10:46:58 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Fri Oct 23 10:54:13 2015 +0200 cleanup: drop debug --- daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c | 1 - 1 files changed, 0 insertions(+), 1 deletions(-) diff --git a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c index aaf9a49..cda19a3 100644 --- a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c +++ b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c @@ -100,7 +100,6 @@ void dmeventd_lvm2_exit(void) lvm2_run(_lvm_handle, "_memlock_dec"); dm_pool_destroy(_mem_pool); _mem_pool = NULL; - log_debug("lvm plugin exiting."); lvm2_exit(_lvm_handle); _lvm_handle = NULL; log_debug("lvm plugin exited.");

1 0

master - cleanup: fix gcc compile with older pthread
by Zdenek Kabelac 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=b5b2a5483450dff1… Commit: b5b2a5483450dff10fd23a7b712649729e619b0d Parent: 21748a86309443ddaefe1fcc0644f0b9a6ea138e Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 23 09:44:21 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Fri Oct 23 10:54:13 2015 +0200 cleanup: fix gcc compile with older pthread Older pthread library was missing 'trick' in pthread_cleanup_pop() which lead to compilation error: error: label at end of compound statement Use explicit ';' to fix it. --- daemons/dmeventd/dmeventd.c | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 949fcf4..b5f1be3 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -1029,6 +1029,9 @@ static void *_monitor_thread(void *arg) } } out: + /* ';' fixes gcc compilation problem with older pthread macros + * "label at end of compound statement" */ + ; pthread_cleanup_pop(1);

1 0

master - cleanup: gcc warning for old-style
by Zdenek Kabelac 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=21748a86309443dd… Commit: 21748a86309443ddaefe1fcc0644f0b9a6ea138e Parent: e5b686d6932be9931a386d15e8d452153225952b Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 23 09:42:38 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Fri Oct 23 10:54:13 2015 +0200 cleanup: gcc warning for old-style --- lib/mm/memlock.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/lib/mm/memlock.c b/lib/mm/memlock.c index 34a6eff..c17ae13 100644 --- a/lib/mm/memlock.c +++ b/lib/mm/memlock.c @@ -639,7 +639,7 @@ void memlock_unlock(struct cmd_context *cmd) #endif -int memlock_count_daemon() +int memlock_count_daemon(void) { return _memlock_count_daemon; }

1 0

master - tools: update for lvm2_disable_dmeventd_monitoring
by Zdenek Kabelac 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=e5b686d6932be993… Commit: e5b686d6932be9931a386d15e8d452153225952b Parent: 87a39d8bac572b67f927dad41548155606a50f4f Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 23 10:48:01 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Fri Oct 23 10:54:13 2015 +0200 tools: update for lvm2_disable_dmeventd_monitoring Make lvm2_disable_dmeventd_monitoring() more explicit. As memlock_inc_daemon() is also used by clvmd, which does changes dmeventd and suspend ignore state at some stages - make updates of these 2 variable tied to the call of lvm2_disable_dmeventd_monitoring(). Once this call is made dmeventd monitoring and suspended devices are ignored. TODO: all lvm-global settings should really be moved to command context. --- lib/misc/lvm-globals.c | 11 +++++++++-- lib/misc/lvm-globals.h | 1 + tools/lvmcmdlib.c | 4 ++-- 3 files changed, 12 insertions(+), 4 deletions(-) diff --git a/lib/misc/lvm-globals.c b/lib/misc/lvm-globals.c index 0cdcc88..06c36e2 100644 --- a/lib/misc/lvm-globals.c +++ b/lib/misc/lvm-globals.c @@ -40,6 +40,8 @@ static int _security_level = SECURITY_LEVEL; static char _cmd_name[30] = ""; static int _mirror_in_sync = 0; static int _dmeventd_monitor = DEFAULT_DMEVENTD_MONITOR; +/* When set, disables update of _dmeventd_monitor & _ignore_suspended_devices */ +static int _disable_dmeventd_monitoring = 0; static int _background_polling = DEFAULT_BACKGROUND_POLLING; static int _ignore_suspended_devices = 0; static int _ignore_lvm_mirrors = DEFAULT_IGNORE_LVM_MIRRORS; @@ -123,10 +125,15 @@ void init_mirror_in_sync(int in_sync) void init_dmeventd_monitor(int reg) { - if (!memlock_count_daemon()) + if (!_disable_dmeventd_monitoring) _dmeventd_monitor = reg; } +void init_disable_dmeventd_monitoring(int reg) +{ + _disable_dmeventd_monitoring = reg; +} + void init_background_polling(int polling) { _background_polling = polling; @@ -134,7 +141,7 @@ void init_background_polling(int polling) void init_ignore_suspended_devices(int ignore) { - if (!memlock_count_daemon()) + if (!_disable_dmeventd_monitoring) _ignore_suspended_devices = ignore; } diff --git a/lib/misc/lvm-globals.h b/lib/misc/lvm-globals.h index 8628a89..132af96 100644 --- a/lib/misc/lvm-globals.h +++ b/lib/misc/lvm-globals.h @@ -38,6 +38,7 @@ void init_lockingfailed(int level); void init_security_level(int level); void init_mirror_in_sync(int in_sync); void init_dmeventd_monitor(int reg); +void init_disable_dmeventd_monitoring(int disable); void init_background_polling(int polling); void init_ignore_suspended_devices(int ignore); void init_ignore_lvm_mirrors(int scan); diff --git a/tools/lvmcmdlib.c b/tools/lvmcmdlib.c index ae8ff5c..b30cce0 100644 --- a/tools/lvmcmdlib.c +++ b/tools/lvmcmdlib.c @@ -80,8 +80,6 @@ int lvm2_run(void *handle, const char *cmdline) /* FIXME Temporary - move to libdevmapper */ ret = ECMD_PROCESSED; if (!strcmp(cmdline, "_memlock_inc")) { - init_dmeventd_monitor(DMEVENTD_MONITOR_IGNORE); - init_ignore_suspended_devices(1); memlock_inc_daemon(cmd); } else if (!strcmp(cmdline, "_memlock_dec")) memlock_dec_daemon(cmd); @@ -99,6 +97,8 @@ int lvm2_run(void *handle, const char *cmdline) void lvm2_disable_dmeventd_monitoring(void *handle) { init_dmeventd_monitor(DMEVENTD_MONITOR_IGNORE); + init_ignore_suspended_devices(1); + init_disable_dmeventd_monitoring(1); /* Lock settings */ } void lvm2_log_level(void *handle, int level)

1 0

master - dmeventd: ensure filter is empty
by Zdenek Kabelac 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=87a39d8bac572b67… Commit: 87a39d8bac572b67f927dad41548155606a50f4f Parent: cff1c728d8c80b3393c461f6229066f633b67fc2 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 23 09:43:21 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Fri Oct 23 10:54:13 2015 +0200 dmeventd: ensure filter is empty On some error path we could have left filter set to some value. Not a big issue - but lets make reporting correct as soon as we can. --- daemons/dmeventd/dmeventd.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index af03091..949fcf4 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -936,6 +936,7 @@ static void _monitor_unregister(void *arg) break; } + thread->events = 0; /* Filter is now empty */ thread->pending = 0; /* Event pending resolved */ thread->processing = 1; /* Process unregistering */

1 0

master - dmeventd: debug error paths
by Zdenek Kabelac 23 Oct '15

23 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=cff1c728d8c80b33… Commit: cff1c728d8c80b3393c461f6229066f633b67fc2 Parent: 2786cd27da3e7c6176940495ba57653ead7cfe08 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 23 09:43:06 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Fri Oct 23 10:54:13 2015 +0200 dmeventd: debug error paths --- daemons/dmeventd/dmeventd.c | 13 +++++++++---- 1 files changed, 9 insertions(+), 4 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 6f3bfff..af03091 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -446,8 +446,10 @@ static int _pthread_create_smallstack(pthread_t *t, void *(*fun)(void *), void * * Linux these functions always succeed (but portable and future-proof * applications should nevertheless handle a possible error return). */ - if ((r = pthread_attr_init(&attr)) != 0) + if ((r = pthread_attr_init(&attr)) != 0) { + log_sys_error("pthread_attr_init", ""); return r; + } /* * We use a smaller stack since it gets preallocated in its entirety @@ -462,7 +464,8 @@ static int _pthread_create_smallstack(pthread_t *t, void *(*fun)(void *), void * t = &tmp; } - r = pthread_create(t, &attr, fun, arg); + if ((r = pthread_create(t, &attr, fun, arg))) + log_sys_error("pthread_create", ""); pthread_attr_destroy(&attr); @@ -1150,7 +1153,7 @@ static int _register_for_event(struct message_data *message_data) } if ((ret = _create_thread(thread))) { - log_sys_error("pthread_create", ""); + stack; _free_thread_status(thread); return -ret; } @@ -1169,8 +1172,10 @@ static int _register_for_event(struct message_data *message_data) usually means we are so starved on resources that we are almost as good as dead already... */ if ((message_data->events_field & DM_EVENT_TIMEOUT) && - (ret = _register_for_timeout(thread))) + (ret = _register_for_timeout(thread))) { + stack; _unregister_for_event(message_data); + } return -ret; }

1 0

master - tests: enable raid test on 4.3
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=2786cd27da3e7c61… Commit: 2786cd27da3e7c6176940495ba57653ead7cfe08 Parent: 1a2d0a0c729e80e4cd122532829f3a11e6566823 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 16:51:18 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:46:10 2015 +0200 tests: enable raid test on 4.3 With kernel -rc6 it's passing our raid tests again - so enable it in tests. Note: related only to Fedora rawhide kernels... --- test/lib/aux.sh | 3 +-- 1 files changed, 1 insertions(+), 2 deletions(-) diff --git a/test/lib/aux.sh b/test/lib/aux.sh index 8a25d87..3f99357 100644 --- a/test/lib/aux.sh +++ b/test/lib/aux.sh @@ -1112,9 +1112,8 @@ have_raid() { # some kernels have broken mdraid bitmaps, don't use them! # may oops kernel, we know for sure all FC24 are currently broken # in general any 4.1, 4.2 is likely useless unless patched - # hopefully 4.3 will be patched case "$(uname -r)" in - 4.[123].*fc24*) return 1 ;; + 4.[12].*fc24*) return 1 ;; esac }

1 0

master - cleanup: indents
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=1a2d0a0c729e80e4… Commit: 1a2d0a0c729e80e4cd122532829f3a11e6566823 Parent: b1319e04020c3cb77b3a5c5a765606efa0b34b83 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 09:57:44 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:46:10 2015 +0200 cleanup: indents --- daemons/dmeventd/dmeventd.c | 8 ++++---- daemons/dmeventd/plugins/raid/dmeventd_raid.c | 4 ++-- lib/metadata/lv_manip.c | 2 +- 3 files changed, 7 insertions(+), 7 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 9662c0e..6f3bfff 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -230,6 +230,7 @@ struct thread_status { void *dso_private; /* dso per-thread status variable */ /* TODO per-thread mutex */ }; + static DM_LIST_INIT(_thread_registry); static DM_LIST_INIT(_thread_registry_unused); @@ -819,6 +820,7 @@ static sigset_t _unblock_sigalrm(void) sigemptyset(&set); sigaddset(&set, SIGALRM); pthread_sigmask(SIG_UNBLOCK, &set, &old); + return old; } @@ -1412,7 +1414,7 @@ fail: * and a complete message is read. Must not block indefinitely. */ static int _client_read(struct dm_event_fifos *fifos, - struct dm_event_daemon_message *msg) + struct dm_event_daemon_message *msg) { struct timeval t; unsigned bytes = 0; @@ -1859,7 +1861,6 @@ static void _daemonize(void) case -1: log_sys_error("fork", ""); exit(EXIT_FAILURE); - case 0: /* Child */ break; @@ -2107,8 +2108,6 @@ int main(int argc, char *argv[]) .server_path = DM_EVENT_FIFO_SERVER }; time_t now, idle_exit_timeout = DMEVENTD_IDLE_EXIT_TIMEOUT; - //struct sys_log logdata = {DAEMON_NAME, LOG_DAEMON}; - opterr = 0; optind = 0; @@ -2195,6 +2194,7 @@ int main(int argc, char *argv[]) /* Signal parent, letting them know we are ready to go. */ if (!_foreground) kill(getppid(), SIGTERM); + log_notice("dmeventd ready for processing."); _idle_since = time(NULL); diff --git a/daemons/dmeventd/plugins/raid/dmeventd_raid.c b/daemons/dmeventd/plugins/raid/dmeventd_raid.c index 2f200ba..99f2680 100644 --- a/daemons/dmeventd/plugins/raid/dmeventd_raid.c +++ b/daemons/dmeventd/plugins/raid/dmeventd_raid.c @@ -42,8 +42,8 @@ static int _process_raid_event(struct dso_state *state, char *params, const char goto out; /* already reported */ log_error("Device #%d of %s array, %s, has failed.", - (int)(d - status->dev_health), - status->raid_type, device); + (int)(d - status->dev_health), + status->raid_type, device); state->failed = 1; if (!dmeventd_lvm2_run_with_lock(state->cmd_lvscan)) diff --git a/lib/metadata/lv_manip.c b/lib/metadata/lv_manip.c index 316710b..738343f 100644 --- a/lib/metadata/lv_manip.c +++ b/lib/metadata/lv_manip.c @@ -4495,7 +4495,7 @@ static int _adjust_policy_params(struct cmd_context *cmd, if (!lv_snapshot_percent(lv, &percent)) return_0; if (!(DM_PERCENT_0 < percent && percent <= DM_PERCENT_100) || percent <= policy_threshold) - return 1; /* nothing to do */ + return 1; /* nothing to do */ } lp->extents = policy_amount;

1 0

master - cleanup: drop uneeded header file
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=b1319e04020c3cb7… Commit: b1319e04020c3cb77b3a5c5a765606efa0b34b83 Parent: 8be60e6a65baf87b12862e07d24bd794608df2f2 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 10:40:51 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:46:10 2015 +0200 cleanup: drop uneeded header file --- lib/locking/no_locking.c | 2 -- 1 files changed, 0 insertions(+), 2 deletions(-) diff --git a/lib/locking/no_locking.c b/lib/locking/no_locking.c index b1dcb30..613877a 100644 --- a/lib/locking/no_locking.c +++ b/lib/locking/no_locking.c @@ -19,8 +19,6 @@ #include "lvm-string.h" #include "activate.h" -#include <signal.h> - /* * No locking */

1 0

master - cleanup: easier to read code
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=8be60e6a65baf87b… Commit: 8be60e6a65baf87b12862e07d24bd794608df2f2 Parent: 39cffa4e9b027162430ea7f3c288256b232a1ed7 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 10:57:57 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:46:10 2015 +0200 cleanup: easier to read code Avoid using #ifdef code and use 'cmd' instead of 'die'. --- daemons/dmeventd/dmeventd.c | 14 ++++---------- 1 files changed, 4 insertions(+), 10 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 606ef30..9662c0e 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -1590,11 +1590,8 @@ static int _do_process_request(struct dm_event_daemon_message *msg) /* Only one caller at a time. */ static void _process_request(struct dm_event_fifos *fifos) { - int die; struct dm_event_daemon_message msg = { 0 }; -#ifdef DEBUG int cmd; -#endif /* * Read the request from the client (client_read, client_write * give true on success and false on failure). @@ -1602,12 +1599,9 @@ static void _process_request(struct dm_event_fifos *fifos) if (!_client_read(fifos, &msg)) return; -#ifdef DEBUG cmd = msg.cmd; -#endif - DEBUGLOG(">>> CMD:%s (0x%x) processing...", decode_cmd(cmd), cmd); - die = (msg.cmd == DM_EVENT_CMD_DIE) ? 1 : 0; + DEBUGLOG(">>> CMD:%s (0x%x) processing...", decode_cmd(cmd), cmd); /* _do_process_request fills in msg (if memory allows for data, otherwise just cmd and size = 0) */ @@ -1616,11 +1610,11 @@ static void _process_request(struct dm_event_fifos *fifos) if (!_client_write(fifos, &msg)) stack; - dm_free(msg.data); + DEBUGLOG("<<< CMD:%s (0x%x) completed (result %d).", decode_cmd(cmd), cmd, msg.cmd); - DEBUGLOG("<<< CMD:%s (0x%x) completed (=%d).", decode_cmd(cmd), cmd, msg.cmd); + dm_free(msg.data); - if (die) { + if (cmd == DM_EVENT_CMD_DIE) { if (unlink(DMEVENTD_PIDFILE)) log_sys_error("unlink", DMEVENTD_PIDFILE); _exit(0);

1 0

master - cleanup: declare vars before code
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=39cffa4e9b027162… Commit: 39cffa4e9b027162430ea7f3c288256b232a1ed7 Parent: 2af696c32f99b2951bcc7e2a73130e46dc4f383f Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 10:33:01 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:46:10 2015 +0200 cleanup: declare vars before code --- libdm/libdm-config.c | 9 +++------ 1 files changed, 3 insertions(+), 6 deletions(-) diff --git a/libdm/libdm-config.c b/libdm/libdm-config.c index b2082c8..c8984b2 100644 --- a/libdm/libdm-config.c +++ b/libdm/libdm-config.c @@ -157,21 +157,18 @@ struct dm_config_tree *dm_config_insert_cascaded_tree(struct dm_config_tree *fir static struct dm_config_node *_config_reverse(struct dm_config_node *head) { - if (!head) - return NULL; - struct dm_config_node *left = head, *middle = NULL, *right = NULL; - do { + while (left) { right = middle; middle = left; left = left->sib; middle->sib = right; middle->child = _config_reverse(middle->child); - } while (left); + } return middle; -}; +} int dm_config_parse(struct dm_config_tree *cft, const char *start, const char *end) {

1 0

master - dmeventd: exit on idle
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=2af696c32f99b295… Commit: 2af696c32f99b2951bcc7e2a73130e46dc4f383f Parent: 4284ba65ebe6401b2dbedc9abe850b650ed68f93 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 20:36:46 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:43:03 2015 +0200 dmeventd: exit on idle Implementing exit when 'dmeventd' is idle. Default idle timeout set to 1 hour - after this time period dmeventd will cleanly exit. On systems with 'systemd' - service is automatically started with next contact on dmeventd communication socket/fifo. On other systems - new dmeventd starts again when lvm2 command detects its missing and monitoring is needed. --- WHATS_NEW_DM | 1 + daemons/dmeventd/dmeventd.c | 35 +++++++++++++++++++++++++++-------- 2 files changed, 28 insertions(+), 8 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 8524a6d..910f6d6 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Support exit on idle for dmenventd (1 hour). Add support to allow unmonitor device from plugin itself. New design for thread co-operation in dmeventd. Dmeventd read device status with 'noflush'. diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index a7c1c6e..606ef30 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -87,11 +87,15 @@ static pthread_mutex_t _global_mutex; static const size_t THREAD_STACK_SIZE = 300 * 1024; +/* Default idle exit timeout 1 hour (in seconds) */ +static const time_t DMEVENTD_IDLE_EXIT_TIMEOUT = 60 * 60; + static int _debug_level = 0; static int _use_syslog = 1; static int _systemd_activation = 0; static int _foreground = 0; static int _restart = 0; +static time_t _idle_since = 0; static char **_initial_registrations = 0; /* FIXME Make configurable at runtime */ @@ -279,6 +283,7 @@ static void _lib_put(struct dso_data *data) DEBUGLOG("Unholding control device."); dm_hold_control_dev(0); dm_lib_release(); + _idle_since = time(NULL); } } } @@ -346,6 +351,7 @@ static struct dso_data *_load_dso(struct message_data *data) if (dm_list_empty(&_dso_registry)) { DEBUGLOG("Holding control device open."); dm_hold_control_dev(1); + _idle_since = 0; } /* @@ -2106,7 +2112,7 @@ int main(int argc, char *argv[]) .client_path = DM_EVENT_FIFO_CLIENT, .server_path = DM_EVENT_FIFO_SERVER }; - int nothreads; + time_t now, idle_exit_timeout = DMEVENTD_IDLE_EXIT_TIMEOUT; //struct sys_log logdata = {DAEMON_NAME, LOG_DAEMON}; opterr = 0; @@ -2197,23 +2203,36 @@ int main(int argc, char *argv[]) kill(getppid(), SIGTERM); log_notice("dmeventd ready for processing."); + _idle_since = time(NULL); + if (_initial_registrations) _process_initial_registrations(); for (;;) { - if (_exit_now) { + if (_idle_since) { + if (_exit_now) { + log_info("dmeventd detected break while being idle " + "for %ld second(s), exiting.", + (long) (time(NULL) - _idle_since)); + break; + } else if (idle_exit_timeout) { + now = time(NULL); + if (now < _idle_since) + _idle_since = now; /* clock change? */ + now -= _idle_since; + if (now >= idle_exit_timeout) { + log_info("dmeventd was idle for %ld second(s), " + "exiting.", (long) now); + break; + } + } + } else if (_exit_now) { _exit_now = 0; /* * When '_exit_now' is set, signal has been received, * but can not simply exit unless all * threads are done processing. */ - _lock_mutex(); - nothreads = (dm_list_empty(&_thread_registry) && - dm_list_empty(&_thread_registry_unused)); - _unlock_mutex(); - if (nothreads) - break; log_warn("WARNING: There are still devices being monitored."); log_warn("WARNING: Refusing to exit."); }

1 0

master - dmeventd: debug signals
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=4284ba65ebe6401b… Commit: 4284ba65ebe6401b2dbedc9abe850b650ed68f93 Parent: aeec62ad1987a5ec59b782f3377d6ae1693e1eab Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 11:15:53 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:41:12 2015 +0200 dmeventd: debug signals --- daemons/dmeventd/dmeventd.c | 20 ++++++++++++++++++++ 1 files changed, 20 insertions(+), 0 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 2706ad6..a7c1c6e 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -789,6 +789,23 @@ static void _unregister_for_timeout(struct thread_status *thread) pthread_mutex_unlock(&_timeout_mutex); } +#ifdef DEBUG_SIGNALS +/* Print list of signals within a signal set */ +static void _print_sigset(const char *prefix, const sigset_t *sigset) +{ + int sig, cnt = 0; + + for (sig = 1; sig < NSIG; sig++) + if (!sigismember(sigset, sig)) { + cnt++; + log_debug("%s%d (%s)", prefix, sig, strsignal(sig)); + } + + if (!cnt) + log_debug("%s<empty signal set>", prefix); +} +#endif + static sigset_t _unblock_sigalrm(void) { sigset_t set, old; @@ -844,6 +861,9 @@ static int _event_wait(struct thread_status *thread) pthread_sigmask(SIG_SETMASK, &set, NULL); +#ifdef DEBUG_SIGNALS + _print_sigset("dmeventd blocking ", &set); +#endif DEBUGLOG("Completed waitevent task for %s.", thread->device.name); return ret;

1 0

master - dmeventd: snapshot plugin unmonitor
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=aeec62ad1987a5ec… Commit: aeec62ad1987a5ec59b782f3377d6ae1693e1eab Parent: 12aa56d29867b962257d7d2789a661a22c649347 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 15:36:21 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:41:12 2015 +0200 dmeventd: snapshot plugin unmonitor Send signal to itself to mark plugin as 'finished' as the watching rule is no longer usable. --- .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 10 ++++++++++ 1 files changed, 10 insertions(+), 0 deletions(-) diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index d9bef5c..7962e96 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -18,6 +18,7 @@ #include <sys/wait.h> #include <stdarg.h> +#include <pthread.h> /* First warning when snapshot is 80% full. */ #define WARNING_THRESH (DM_PERCENT_1 * 80) @@ -203,6 +204,15 @@ void process_event(struct dm_task *dmt, /* Maybe configurable ? */ _remove(dm_task_get_uuid(dmt)); #endif + pthread_kill(pthread_self(), SIGALRM); + goto out; + } + + if (length <= (status->used_sectors - status->metadata_sectors)) { + /* TODO eventually recognize earlier when room is enough */ + log_info("Dropping monitoring of fully provisioned snapshot %s.", + device); + pthread_kill(pthread_self(), SIGALRM); goto out; }

1 0

master - dmeventd: handle signal from plugin
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=12aa56d29867b962… Commit: 12aa56d29867b962257d7d2789a661a22c649347 Parent: 9156c5d0888bf95b79d931682b51fc63c96ba236 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 15:48:14 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:40:07 2015 +0200 dmeventd: handle signal from plugin Add support to unmonitor device when monitor recognizes there is nothing to monitor anymore. TODO: possibly API change with return value could be also used. --- WHATS_NEW_DM | 3 ++- daemons/dmeventd/dmeventd.c | 11 +++++++++++ 2 files changed, 13 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 8c27139..8524a6d 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,6 +1,7 @@ Version 1.02.110 - ====================================== - New design for thread cooperation in dmeventd. + Add support to allow unmonitor device from plugin itself. + New design for thread co-operation in dmeventd. Dmeventd read device status with 'noflush'. Dmeventd closes control device when no device is monitored. Thin plugin for dmeventd improved percentage usage. diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index d8c317e..2706ad6 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -973,6 +973,17 @@ static void *_monitor_thread(void *arg) _lock_mutex(); thread->processing = 0; + + /* + * Thread can terminate itself from plugin via SIGALRM + * Timer thread will not send signal while processing + * TODO: maybe worth API change and return value for + * _do_process_event() instead of this signal solution + */ + if (sigpending(&pendmask) < 0) + log_sys_error("sigpending", ""); + else if (sigismember(&pendmask, SIGALRM)) + break; } else { _unlock_mutex();

1 0

master - dmeventd: rework locking code
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=9156c5d0888bf95b… Commit: 9156c5d0888bf95b79d931682b51fc63c96ba236 Parent: 466a1c72b7f24be4c932b503b3f8a3fb50a2eda5 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 15:47:53 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:39:24 2015 +0200 dmeventd: rework locking code Redesign threading code: - plugin registration runs within its new created thread for improved parallel usage. - wait task is created just once and used during whole plugin lifetime. - event thread is based over 'events' filter being set - when filter is 0, such thread is 'unused'. - event loop is simplified. - timeout thread is never signaling 'processing' thread. - pending of events filter cnange is properly reported and running event thread is signalled when possible. - helgrind is not reporting problems. --- WHATS_NEW_DM | 1 + daemons/dmeventd/dmeventd.c | 497 ++++++++++++++++++++----------------------- 2 files changed, 231 insertions(+), 267 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 5f92657..8c27139 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + New design for thread cooperation in dmeventd. Dmeventd read device status with 'noflush'. Dmeventd closes control device when no device is monitored. Thin plugin for dmeventd improved percentage usage. diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 8267dc4..d8c317e 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -85,23 +85,6 @@ static volatile sig_atomic_t _exit_now = 0; /* set to '1' when signal is given t */ static pthread_mutex_t _global_mutex; -/* - There are three states a thread can attain (see struct - thread_status, field int status): - - - DM_THREAD_RUNNING: thread has started up and is either working or - waiting for events... transitions to either SHUTDOWN or DONE - - DM_THREAD_SHUTDOWN: thread is still doing something, but it is - supposed to terminate (and transition to DONE) as soon as it - finishes whatever it was doing at the point of flipping state to - SHUTDOWN... the thread is still on the thread list - - DM_THREAD_DONE: thread has terminated and has been moved over to - unused thread list, cleanup pending - */ -#define DM_THREAD_RUNNING 0 -#define DM_THREAD_SHUTDOWN 1 -#define DM_THREAD_DONE 2 - static const size_t THREAD_STACK_SIZE = 300 * 1024; static int _debug_level = 0; @@ -204,6 +187,13 @@ struct message_data { struct dm_event_daemon_message *msg; /* Pointer to message buffer. */ }; +/* There are three states a thread can attain. */ +enum { + DM_THREAD_REGISTERING, /* Registering, transitions to RUNNING */ + DM_THREAD_RUNNING, /* Working on events, transitions to DONE */ + DM_THREAD_DONE /* Terminated and cleanup is pending */ +}; + /* * Housekeeping of thread+device states. * @@ -222,18 +212,19 @@ struct thread_status { char *name; int major, minor; } device; - uint32_t event_nr; /* event number */ int processing; /* Set when event is being processed */ - int status; /* see DM_THREAD_{RUNNING,SHUTDOWN,DONE} - constants above */ - enum dm_event_mask events; /* bitfield for event filter. */ - enum dm_event_mask current_events; /* bitfield for occured events. */ - struct dm_task *current_task; + int status; /* See DM_THREAD_{REGISTERING,RUNNING,DONE} */ + + int events; /* bitfield for event filter. */ + int current_events; /* bitfield for occured events. */ + struct dm_task *wait_task; + int pending; /* Set when event filter change is pending */ time_t next_time; uint32_t timeout; struct dm_list timeout_list; void *dso_private; /* dso per-thread status variable */ + /* TODO per-thread mutex */ }; static DM_LIST_INIT(_thread_registry); static DM_LIST_INIT(_thread_registry_unused); @@ -379,34 +370,57 @@ bad: /* Allocate/free the thread status structure for a monitoring thread. */ static void _free_thread_status(struct thread_status *thread) { + _lib_put(thread->dso_data); - if (thread->current_task) - dm_task_destroy(thread->current_task); + if (thread->wait_task) + dm_task_destroy(thread->wait_task); dm_free(thread->device.uuid); dm_free(thread->device.name); dm_free(thread); } -/* Allocate/free the status structure for a monitoring thread. */ +/* Note: events_field must not be 0, ensured by caller */ static struct thread_status *_alloc_thread_status(const struct message_data *data, struct dso_data *dso_data) { - struct thread_status *ret; - - if (!(ret = dm_zalloc(sizeof(*ret)))) - return NULL; + struct thread_status *thread; - if (!(ret->device.uuid = dm_strdup(data->device_uuid))) { - dm_free(ret); + if (!(thread = dm_zalloc(sizeof(*thread)))) { + log_error("Cannot create new thread, out of memory."); return NULL; } - ret->dso_data = dso_data; - ret->events = data->events_field; - ret->timeout = data->timeout_secs; - dm_list_init(&ret->timeout_list); + _lib_get(dso_data); + thread->dso_data = dso_data; - return ret; + if (!(thread->wait_task = dm_task_create(DM_DEVICE_WAITEVENT))) + goto_out; + + if (!dm_task_set_uuid(thread->wait_task, data->device_uuid)) + goto_out; + + if (!(thread->device.uuid = dm_strdup(data->device_uuid))) + goto_out; + + /* Until real name resolved, use UUID */ + if (!(thread->device.name = dm_strdup(data->device_uuid))) + goto_out; + + /* runs ioctl and may register lvm2 pluging */ + thread->processing = 1; + thread->status = DM_THREAD_REGISTERING; + + thread->events = data->events_field; + thread->pending = DM_EVENT_REGISTRATION_PENDING; + thread->timeout = data->timeout_secs; + dm_list_init(&thread->timeout_list); + + return thread; + +out: + _free_thread_status(thread); + + return NULL; } /* @@ -563,6 +577,8 @@ static int _fill_device_data(struct thread_status *ts) ts->device.major = dmi.major; ts->device.minor = dmi.minor; + dm_task_set_event_nr(ts->wait_task, dmi.event_nr); + ret = 1; fail: dm_task_destroy(dmt); @@ -711,8 +727,17 @@ static void *_timeout_thread(void *unused __attribute__((unused))) dm_list_iterate_items_gen(thread, &_timeout_registry, timeout_list) { if (thread->next_time <= curr_time) { thread->next_time = curr_time + thread->timeout; - DEBUGLOG("Sending SIGALRM to Thr %x for timeout.", (int) thread->thread); - pthread_kill(thread->thread, SIGALRM); + _lock_mutex(); + if (thread->processing) { + /* Cannot signal processing monitoring thread */ + log_debug("Skipping SIGALRM to processing Thr %x for timeout.", + (int) thread->thread); + } else { + DEBUGLOG("Sending SIGALRM to Thr %x for timeout.", + (int) thread->thread); + pthread_kill(thread->thread, SIGALRM); + } + _unlock_mutex(); } if (thread->next_time < timeout.tv_sec || !timeout.tv_sec) @@ -781,69 +806,45 @@ enum { }; /* Wait on a device until an event occurs. */ -static int _event_wait(struct thread_status *thread, struct dm_task **task) +static int _event_wait(struct thread_status *thread) { sigset_t set; int ret = DM_WAIT_RETRY; - struct dm_task *dmt; struct dm_info info; - int ioctl_errno; - - *task = 0; - DEBUGLOG("Preparing waitevent task for %s", thread->device.uuid); - if (!(dmt = dm_task_create(DM_DEVICE_WAITEVENT))) - return DM_WAIT_RETRY; + /* TODO: audit libdm thread usage */ - thread->current_task = dmt; - - if (!dm_task_set_uuid(dmt, thread->device.uuid) || - !dm_task_set_event_nr(dmt, thread->event_nr)) - goto out; - - /* - * Check if there are already some waiting events, - * in this case the logging is unmodified. - * TODO: audit libdm thread usage - */ - DEBUGLOG("Starting waitevent task for %s", thread->device.uuid); /* * This is so that you can break out of waiting on an event, * either for a timeout event, or to cancel the thread. */ set = _unblock_sigalrm(); - if (dm_task_run(dmt)) { + + if (dm_task_run(thread->wait_task)) { thread->current_events |= DM_EVENT_DEVICE_ERROR; ret = DM_WAIT_INTR; - - if ((ret = dm_task_get_info(dmt, &info))) - thread->event_nr = info.event_nr; + /* Update event_nr */ + if (dm_task_get_info(thread->wait_task, &info)) + dm_task_set_event_nr(thread->wait_task, info.event_nr); } else { - ioctl_errno = dm_task_get_errno(dmt); - if (thread->events & DM_EVENT_TIMEOUT && ioctl_errno == EINTR) { + switch (dm_task_get_errno(thread->wait_task)) { + case ENXIO: + log_error("%s disappeared, detaching.", + thread->device.name); + ret = DM_WAIT_FATAL; + break; + case EINTR: thread->current_events |= DM_EVENT_TIMEOUT; ret = DM_WAIT_INTR; - } else if (thread->status == DM_THREAD_SHUTDOWN && ioctl_errno == EINTR) - ret = DM_WAIT_FATAL; - else { - if (ioctl_errno == ENXIO) { - log_error("%s disappeared, detaching.", - thread->device.name); - ret = DM_WAIT_FATAL; - } else - log_sys_error("dm_task_run", ""); + break; + default: + log_sys_error("dm_task_run", "waitevent"); } } - DEBUGLOG("Completed waitevent task for %s", thread->device.uuid); pthread_sigmask(SIG_SETMASK, &set, NULL); - out: - if (ret == DM_WAIT_FATAL || ret == DM_WAIT_RETRY) { - dm_task_destroy(dmt); - thread->current_task = NULL; - } else - *task = dmt; + DEBUGLOG("Completed waitevent task for %s.", thread->device.name); return ret; } @@ -869,9 +870,27 @@ static int _do_unregister_device(struct thread_status *thread) } /* Process an event in the DSO. */ -static void _do_process_event(struct thread_status *thread, struct dm_task *task) +static void _do_process_event(struct thread_status *thread) { - thread->dso_data->process_event(task, thread->current_events, &(thread->dso_private)); + struct dm_task *task; + + /* NOTE: timeout event gets status */ + task = (thread->current_events & DM_EVENT_TIMEOUT) + ? _get_device_status(thread) : thread->wait_task; + + if (!task) + log_error("Lost event in Thr %x.", (int)thread->thread); + else { + thread->dso_data->process_event(task, thread->current_events, &(thread->dso_private)); + if (task != thread->wait_task) + dm_task_destroy(task); + } +} + +static void _thread_unused(struct thread_status *thread) +{ + UNLINK_THREAD(thread); + LINK(thread, &_thread_registry_unused); } /* Thread cleanup handler to unregister device. */ @@ -879,36 +898,30 @@ static void _monitor_unregister(void *arg) { struct thread_status *thread = arg, *thread_iter; - DEBUGLOG("_monitor_unregister thread cleanup handler running"); - if (!_do_unregister_device(thread)) + dm_list_iterate_items(thread_iter, &_thread_registry) + if (thread_iter == thread) { + /* Relink to _unused */ + _thread_unused(thread); + break; + } + + thread->pending = 0; /* Event pending resolved */ + thread->processing = 1; /* Process unregistering */ + + _unlock_mutex(); + + DEBUGLOG("Unregistering monitor for %s.", thread->device.name); + _unregister_for_timeout(thread); + + if ((thread->status != DM_THREAD_REGISTERING) && + !_do_unregister_device(thread)) log_error("%s: %s unregister failed.", __func__, thread->device.name); - if (thread->current_task) { - dm_task_destroy(thread->current_task); - thread->current_task = NULL; - } - _lock_mutex(); - if (thread->events & DM_EVENT_TIMEOUT) { - /* _unregister_for_timeout locks another mutex, we - don't want to deadlock so we release our mutex for - a bit */ - _unlock_mutex(); - _unregister_for_timeout(thread); - _lock_mutex(); - } - /* we may have been relinked to unused registry since we were - called, so check that */ - dm_list_iterate_items(thread_iter, &_thread_registry_unused) - if (thread_iter == thread) { - thread->status = DM_THREAD_DONE; - _unlock_mutex(); - return; - } DEBUGLOG("Marking Thr %x as DONE and unused.", (int)thread->thread); - thread->status = DM_THREAD_DONE; - UNLINK_THREAD(thread); - LINK(thread, &_thread_registry_unused); + + _lock_mutex(); + thread->status = DM_THREAD_DONE; /* Last access to thread memory! */ _unlock_mutex(); } @@ -916,80 +929,64 @@ static void _monitor_unregister(void *arg) static void *_monitor_thread(void *arg) { struct thread_status *thread = arg; - int wait_error; - struct dm_task *task; + int ret; + sigset_t pendmask; pthread_setcanceltype(PTHREAD_CANCEL_DEFERRED, NULL); pthread_cleanup_push(_monitor_unregister, thread); - /* Wait for do_process_request() to finish its task. */ - _lock_mutex(); - thread->status = DM_THREAD_RUNNING; - _unlock_mutex(); - - /* Loop forever awaiting/analyzing device events. */ - while (1) { - thread->current_events = 0; + if (!_fill_device_data(thread)) { + log_error("Failed to fill device data for %s.", thread->device.uuid); + _lock_mutex(); + goto out; + } - wait_error = _event_wait(thread, &task); - if (wait_error == DM_WAIT_RETRY) { - usleep(100); /* avoid busy loop */ - continue; - } + if (!_do_register_device(thread)) { + log_error("Failed to register device %s.", thread->device.name); + _lock_mutex(); + goto out; + } - if (wait_error == DM_WAIT_FATAL) - break; + _lock_mutex(); + thread->status = DM_THREAD_RUNNING; + thread->processing = 0; - /* Timeout occurred, task is not filled properly. - * We get device status here for processing it in DSO. - */ - if (wait_error == DM_WAIT_INTR && - thread->current_events & DM_EVENT_TIMEOUT) { - dm_task_destroy(task); - task = _get_device_status(thread); - /* FIXME: syslog fail here ? */ - if (!(thread->current_task = task)) - continue; - } + /* Loop awaiting/analyzing device events. */ + while (thread->events) { - /* - * We know that wait succeeded and stored a - * pointer to dm_task with device status into task. - */ + thread->pending = 0; /* Event is no longer pending... */ /* - * Check against filter. + * Check against bitmask filter. * * If there's current events delivered from _event_wait() AND * the device got registered for those events AND * those events haven't been processed yet, call * the DSO's process_event() handler. */ - _lock_mutex(); - if (thread->status == DM_THREAD_SHUTDOWN) { - _unlock_mutex(); - break; - } - if (thread->events & thread->current_events) { - thread->processing = 1; + thread->processing = 1; /* Cannot be removed/signaled */ _unlock_mutex(); - _do_process_event(thread, task); - dm_task_destroy(task); - thread->current_task = NULL; + _do_process_event(thread); + thread->current_events = 0; /* Current events processed */ _lock_mutex(); thread->processing = 0; - _unlock_mutex(); } else { _unlock_mutex(); - dm_task_destroy(task); - thread->current_task = NULL; + + if ((ret = _event_wait(thread)) == DM_WAIT_RETRY) + usleep(100); /* Avoid busy loop, wait without mutex */ + + _lock_mutex(); + + if (ret == DM_WAIT_FATAL) + break; } } +out: - DEBUGLOG("Finished _monitor_thread."); pthread_cleanup_pop(1); return NULL; @@ -1001,10 +998,36 @@ static int _create_thread(struct thread_status *thread) return _pthread_create_smallstack(&thread->thread, _monitor_thread, thread); } -static int _terminate_thread(struct thread_status *thread) +/* Update events - needs to be locked */ +static int _update_events(struct thread_status *thread, int events) { - DEBUGLOG("Sending SIGALRM to terminate Thr %x.", (int)thread->thread); - return pthread_kill(thread->thread, SIGALRM); + int ret = 0; + + if (thread->events == events) + return 0; /* Nothing has changed */ + + thread->events = events; + thread->pending = DM_EVENT_REGISTRATION_PENDING; + + /* Only non-processing threads can be notified */ + if (!thread->processing) { + DEBUGLOG("Sending SIGALRM to wakeup Thr %x.", (int)thread->thread); + + /* Notify thread waiting in ioctl (to speed-up) */ + if ((ret = pthread_kill(thread->thread, SIGALRM))) { + if (ret == ESRCH) + thread->events = 0; /* thread is gone */ + else + log_error("Unable to wakeup thread: %s", + strerror(ret)); + } + } + + /* Threads with no events has to be moved to unused */ + if (!thread->events) + _thread_unused(thread); + + return -ret; } /* Return success on daemon active check. */ @@ -1020,8 +1043,8 @@ static int _active(struct message_data *message_data) */ static int _unregister_for_event(struct message_data *message_data) { - int ret = 0; struct thread_status *thread; + int ret; /* * Clear event in bitfield and deactivate @@ -1031,38 +1054,21 @@ static int _unregister_for_event(struct message_data *message_data) if (!(thread = _lookup_thread_status(message_data))) { _unlock_mutex(); - ret = -ENODEV; - goto out; + return -ENODEV; } - if (thread->status == DM_THREAD_DONE) { - /* the thread has terminated while we were not - watching */ - _unlock_mutex(); - return 0; - } + /* AND mask event ~# from events bitfield. */ + ret = _update_events(thread, (thread->events & ~message_data->events_field)); - thread->events &= ~message_data->events_field; + _unlock_mutex(); - if (!(thread->events & DM_EVENT_TIMEOUT)) { - _unlock_mutex(); + /* If there are no events, thread is later garbage + * collected by _cleanup_unused_threads */ + if (message_data->events_field & DM_EVENT_TIMEOUT) _unregister_for_timeout(thread); - _lock_mutex(); - } - /* - * In case there's no events to monitor on this device -> - * unlink and terminate its monitoring thread. - */ - if (!thread->events) { - DEBUGLOG("Marking Thr %x unused (no events).", (int)thread->thread); - UNLINK_THREAD(thread); - LINK(thread, &_thread_registry_unused); - } - _unlock_mutex(); - DEBUGLOG("Unregistered uuid:%s.", thread->device.uuid); + DEBUGLOG("Unregistered event for %s.", thread->device.name); - out: return ret; } @@ -1082,73 +1088,52 @@ static int _register_for_event(struct message_data *message_data) !(dso_data = _load_dso(message_data))) { stack; #ifdef ELIBACC - ret = -ELIBACC; + ret = ELIBACC; #else - ret = -ENODEV; + ret = ENODEV; #endif return ret; } _lock_mutex(); - if ((thread = _lookup_thread_status(message_data))) - /* Or event # into events bitfield. */ - thread->events |= message_data->events_field; - - _unlock_mutex(); + if ((thread = _lookup_thread_status(message_data))) { + /* OR event # into events bitfield. */ + ret = _update_events(thread, (thread->events | message_data->events_field)); + } else { + _unlock_mutex(); - if (!thread) { + /* Only creating thread during event processing + * Remaining initialization happens within monitoring thread */ if (!(thread = _alloc_thread_status(message_data, dso_data))) { stack; return -ENOMEM; } - if (!_fill_device_data(thread)) { - ret = -ENODEV; - goto_out; - } - - if (!_do_register_device(thread)) { - ret = -ENOMEM; - goto_out; - } - - if ((ret = -_create_thread(thread))) { - _do_unregister_device(thread); - goto_out; + if ((ret = _create_thread(thread))) { + log_sys_error("pthread_create", ""); + _free_thread_status(thread); + return -ret; } _lock_mutex(); - if (_lookup_thread_status(message_data)) { - DEBUGLOG("Race, uuid already registered, marking Thr %x unused.", - (int)thread->thread); - thread->status = DM_THREAD_SHUTDOWN; - thread->events = 0; - LINK(thread, &_thread_registry_unused); - _unlock_mutex(); - ret = -EEXIST; /* race ? */ - goto_out; - } - + /* Note: same uuid can't be added in parallel */ LINK_THREAD(thread); - _unlock_mutex(); } + _unlock_mutex(); + /* If creation of timeout thread fails (as it may), we fail here completely. The client is responsible for either retrying later or trying to register without timeout events. However, if timeout thread cannot be started, it usually means we are so starved on resources that we are almost as good as dead already... */ - if ((thread->events & DM_EVENT_TIMEOUT) && - (ret = -_register_for_timeout(thread))) + if ((message_data->events_field & DM_EVENT_TIMEOUT) && + (ret = _register_for_timeout(thread))) _unregister_for_event(message_data); - return ret; -out: - _free_thread_status(thread); - - return ret; + return -ret; } /* @@ -1161,16 +1146,14 @@ static int _registered_device(struct message_data *message_data, { int r; struct dm_event_daemon_message *msg = message_data->msg; - unsigned events = ((thread->status == DM_THREAD_RUNNING) && - thread->events) ? thread->events : - thread->events | DM_EVENT_REGISTRATION_PENDING; dm_free(msg->data); if ((r = dm_asprintf(&(msg->data), "%s %s %s %u", message_data->id, thread->dso_data->dso_name, - thread->device.uuid, events)) < 0) + thread->device.uuid, + thread->events | thread->pending)) < 0) return -ENOMEM; msg->size = (uint32_t) r; @@ -1182,7 +1165,6 @@ static int _registered_device(struct message_data *message_data, static int _want_registered_device(char *dso_name, char *device_uuid, struct thread_status *thread) { - DEBUGLOG("Looking for dso:%s uuid:%s.", dso_name, device_uuid); /* If DSO names and device paths are equal. */ if (dso_name && device_uuid) return !strcmp(dso_name, thread->dso_data->dso_name) && @@ -1625,58 +1607,39 @@ static void _process_initial_registrations(void) static void _cleanup_unused_threads(void) { - int ret; struct dm_list *l; struct thread_status *thread; - int join_ret = 0; + int ret; _lock_mutex(); + while ((l = dm_list_first(&_thread_registry_unused))) { thread = dm_list_item(l, struct thread_status); - if (thread->processing) - break; /* cleanup on the next round */ + if (thread->status != DM_THREAD_DONE) { + if (thread->processing) + break; /* cleanup on the next round */ - if (thread->status == DM_THREAD_RUNNING) { - thread->status = DM_THREAD_SHUTDOWN; - break; - } + /* Signal possibly sleeping thread */ + ret = pthread_kill(thread->thread, SIGALRM); + if (!ret || (ret != ESRCH)) + break; /* check again on the next round */ - if (thread->status == DM_THREAD_SHUTDOWN) { - if (!thread->events) { - /* turn codes negative -- should we be returning this? */ - ret = _terminate_thread(thread); + /* thread is likely gone */ + } - if (ret == ESRCH) { - thread->status = DM_THREAD_DONE; - } else if (ret) { - log_error("Unable to terminate thread: %s", - strerror(ret)); - } - break; - } + dm_list_del(l); + _unlock_mutex(); - dm_list_del(l); - log_error("thread can't be on unused list unless !thread->events"); - thread->status = DM_THREAD_RUNNING; - LINK_THREAD(thread); + DEBUGLOG("Destroying Thr %x.", (int)thread->thread); - continue; - } + if (pthread_join(thread->thread, NULL)) + log_sys_error("pthread_join", ""); - if (thread->status == DM_THREAD_DONE) { - dm_list_del(l); - _unlock_mutex(); - DEBUGLOG("Destroying Thr %x.", (int)thread->thread); - join_ret = pthread_join(thread->thread, NULL); - _free_thread_status(thread); - _lock_mutex(); - } + _free_thread_status(thread); + _lock_mutex(); } _unlock_mutex(); - - if (join_ret) - log_error("Failed pthread_join: %s.", strerror(join_ret)); } static void _sig_alarm(int signum __attribute__((unused)))

1 0

master - cleanup: use enums
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=466a1c72b7f24be4… Commit: 466a1c72b7f24be4c932b503b3f8a3fb50a2eda5 Parent: 81e9ab3156badecc6a64447708c4ae4886e3c244 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Wed Oct 21 20:56:20 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:36:00 2015 +0200 cleanup: use enums --- daemons/dmeventd/dmeventd.c | 10 ++++++---- 1 files changed, 6 insertions(+), 4 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index cde4f93..8267dc4 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -102,7 +102,7 @@ static pthread_mutex_t _global_mutex; #define DM_THREAD_SHUTDOWN 1 #define DM_THREAD_DONE 2 -#define THREAD_STACK_SIZE (300*1024) +static const size_t THREAD_STACK_SIZE = 300 * 1024; static int _debug_level = 0; static int _use_syslog = 1; @@ -774,9 +774,11 @@ static sigset_t _unblock_sigalrm(void) return old; } -#define DM_WAIT_RETRY 0 -#define DM_WAIT_INTR 1 -#define DM_WAIT_FATAL 2 +enum { + DM_WAIT_RETRY, + DM_WAIT_INTR, + DM_WAIT_FATAL +}; /* Wait on a device until an event occurs. */ static int _event_wait(struct thread_status *thread, struct dm_task **task)

1 0

master - dmeventd: code mode _get_device_status
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=81e9ab3156badecc… Commit: 81e9ab3156badecc6a64447708c4ae4886e3c244 Parent: 15dbd4b56a65dba2f19bb1685529731faa9894c6 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 12:36:25 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:35:25 2015 +0200 dmeventd: code mode _get_device_status Move _get_device_status() in code. Use dm_task_no_flush() function when reading status. (e.g. none blocking for thins pool) --- WHATS_NEW_DM | 1 + daemons/dmeventd/dmeventd.c | 44 +++++++++++++++++++++++------------------- 2 files changed, 25 insertions(+), 20 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 000bddd..5f92657 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Dmeventd read device status with 'noflush'. Dmeventd closes control device when no device is monitored. Thin plugin for dmeventd improved percentage usage. Snapshot plugin for dmeventd improved percentage usage. diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 26e611d..cde4f93 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -570,6 +570,30 @@ fail: return ret; } +static struct dm_task *_get_device_status(struct thread_status *ts) +{ + struct dm_task *dmt = dm_task_create(DM_DEVICE_STATUS); + + if (!dmt) + return_NULL; + + if (!dm_task_set_uuid(dmt, ts->device.uuid)) { + dm_task_destroy(dmt); + return_NULL; + } + + /* Non-blocking status read */ + if (!dm_task_no_flush(dmt)) + log_warn("WARNING: Can't set no_flush for dm status."); + + if (!dm_task_run(dmt)) { + dm_task_destroy(dmt); + return_NULL; + } + + return dmt; +} + /* * Find an existing thread for a device. * @@ -886,26 +910,6 @@ static void _monitor_unregister(void *arg) _unlock_mutex(); } -static struct dm_task *_get_device_status(struct thread_status *ts) -{ - struct dm_task *dmt = dm_task_create(DM_DEVICE_STATUS); - - if (!dmt) - return NULL; - - if (!dm_task_set_uuid(dmt, ts->device.uuid)) { - dm_task_destroy(dmt); - return NULL; - } - - if (!dm_task_run(dmt)) { - dm_task_destroy(dmt); - return NULL; - } - - return dmt; -} - /* Device monitoring thread. */ static void *_monitor_thread(void *arg) {

1 0

master - dmeventd: minimize locking time for get_imeout
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=15dbd4b56a65dba2… Commit: 15dbd4b56a65dba2f19bb1685529731faa9894c6 Parent: e2ea2a81476495f494915774c46c0b3c139d1c3f Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 12:38:26 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:34:30 2015 +0200 dmeventd: minimize locking time for get_imeout Don't hold lock when creating message (allocating memory). Thread cannot dissapear as it's only the same thread which may clean it. --- daemons/dmeventd/dmeventd.c | 18 +++++++++--------- 1 files changed, 9 insertions(+), 9 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 2f5fe67..26e611d 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -1292,18 +1292,18 @@ static int _get_timeout(struct message_data *message_data) struct thread_status *thread; struct dm_event_daemon_message *msg = message_data->msg; - dm_free(msg->data); - _lock_mutex(); - if ((thread = _lookup_thread_status(message_data))) { - msg->size = dm_asprintf(&(msg->data), "%s %" PRIu32, - message_data->id, thread->timeout); - } else - msg->data = NULL; - + thread = _lookup_thread_status(message_data); _unlock_mutex(); - return thread ? 0 : -ENODEV; + if (!thread) + return -ENODEV; + + dm_free(msg->data); + msg->size = dm_asprintf(&(msg->data), "%s %" PRIu32, + message_data->id, thread->timeout); + + return (msg->data && msg->size) ? 0 : -ENOMEM; } /* Open fifos used for client communication. */

1 0

master - dmeventd: drop unneded test
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=e2ea2a81476495f4… Commit: e2ea2a81476495f494915774c46c0b3c139d1c3f Parent: 941c6354dbf783da17851be7e38f243334a4acd6 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 12:31:45 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:34:29 2015 +0200 dmeventd: drop unneded test Function is never called without device.uuid. --- daemons/dmeventd/dmeventd.c | 3 --- 1 files changed, 0 insertions(+), 3 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index bbdee82..2f5fe67 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -545,9 +545,6 @@ static int _fill_device_data(struct thread_status *ts) struct dm_info dmi; int ret = 0; - if (!ts->device.uuid) - return 0; - if (!(dmt = dm_task_create(DM_DEVICE_INFO))) return 0;

1 0

master - dmeventd: wake up timer when setting new timeout
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=941c6354dbf783da… Commit: 941c6354dbf783da17851be7e38f243334a4acd6 Parent: 02eb000f5145aa60578b4591053bff9a797e64cc Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 12:37:18 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:34:29 2015 +0200 dmeventd: wake up timer when setting new timeout --- daemons/dmeventd/dmeventd.c | 15 ++++++++++++--- 1 files changed, 12 insertions(+), 3 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 885aa9d..bbdee82 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -1274,11 +1274,20 @@ static int _set_timeout(struct message_data *message_data) struct thread_status *thread; _lock_mutex(); - if ((thread = _lookup_thread_status(message_data))) - thread->timeout = message_data->timeout_secs; + thread = _lookup_thread_status(message_data); _unlock_mutex(); - return thread ? 0 : -ENODEV; + if (!thread) + return -ENODEV; + + /* Lets reprogram timer */ + pthread_mutex_lock(&_timeout_mutex); + thread->timeout = message_data->timeout_secs; + thread->next_time = 0; + pthread_cond_signal(&_timeout_cond); + pthread_mutex_unlock(&_timeout_mutex); + + return 0; } static int _get_timeout(struct message_data *message_data)

1 0

master - dmeventd: use dm_hold_control_dev
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=02eb000f5145aa60… Commit: 02eb000f5145aa60578b4591053bff9a797e64cc Parent: efc76ca33dec99e3b0499ef28fb9236815ad2258 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Wed Oct 21 20:52:29 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:34:27 2015 +0200 dmeventd: use dm_hold_control_dev Need here to keep control device opened while there is 'any' dso plugin loaded - otherwise there would a race closing controlfd inside lvm2 plugin while some other monitoring thread would tried to execute another WAITEVENT task. --- WHATS_NEW_DM | 1 + daemons/dmeventd/dmeventd.c | 13 +++++++++++++ 2 files changed, 14 insertions(+), 0 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 2b87e4b..000bddd 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Dmeventd closes control device when no device is monitored. Thin plugin for dmeventd improved percentage usage. Snapshot plugin for dmeventd improved percentage usage. Add dm_hold_control_dev to allow holding of control device open. diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 69fe393..885aa9d 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -282,6 +282,13 @@ static void _lib_put(struct dso_data *data) dlclose(data->dso_handle); UNLINK_DSO(data); _free_dso_data(data); + + /* Close control device if there is no plugin in-use */ + if (dm_list_empty(&_dso_registry)) { + DEBUGLOG("Unholding control device."); + dm_hold_control_dev(0); + dm_lib_release(); + } } } @@ -344,6 +351,12 @@ static struct dso_data *_load_dso(struct message_data *data) goto_bad; } + /* Keep control device open until last user closes */ + if (dm_list_empty(&_dso_registry)) { + DEBUGLOG("Holding control device open."); + dm_hold_control_dev(1); + } + /* * Keep handle to close the library once * we've got no references to it any more.

1 0

master - dmeventd: move dso handling into single code section
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=efc76ca33dec99e3… Commit: efc76ca33dec99e3b0499ef28fb9236815ad2258 Parent: 590091a4faa4a4eb66598137576e01e692fb9f8d Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 12:22:55 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:33:19 2015 +0200 dmeventd: move dso handling into single code section Move all DSO related function in front, so they could be easily referenced from rest of code. Add proper error paths with logging and error reporting. Drop mutex locking when releasing DSO - since DSO is always allocated and released in main 'event' processing thread. --- daemons/dmeventd/dmeventd.c | 236 ++++++++++++++++++++++--------------------- 1 files changed, 122 insertions(+), 114 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 93df32e..69fe393 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -243,29 +243,127 @@ static DM_LIST_INIT(_timeout_registry); static pthread_mutex_t _timeout_mutex = PTHREAD_MUTEX_INITIALIZER; static pthread_cond_t _timeout_cond = PTHREAD_COND_INITIALIZER; -/* Allocate/free the status structure for a monitoring thread. */ -static struct thread_status *_alloc_thread_status(const struct message_data *data, - struct dso_data *dso_data) + +/********** + * DSO + **********/ + +/* DSO data allocate/free. */ +static void _free_dso_data(struct dso_data *data) { - struct thread_status *ret; + dm_free(data->dso_name); + dm_free(data); +} - if (!(ret = dm_zalloc(sizeof(*ret)))) - return NULL; +static struct dso_data *_alloc_dso_data(struct message_data *data) +{ + struct dso_data *ret = (typeof(ret)) dm_zalloc(sizeof(*ret)); - if (!(ret->device.uuid = dm_strdup(data->device_uuid))) { + if (!ret) + return_NULL; + + if (!(ret->dso_name = dm_strdup(data->dso_name))) { dm_free(ret); - return NULL; + return_NULL; } - ret->dso_data = dso_data; - ret->events = data->events_field; - ret->timeout = data->timeout_secs; - dm_list_init(&ret->timeout_list); + return ret; +} + +/* DSO reference counting. */ +static void _lib_get(struct dso_data *data) +{ + data->ref_count++; +} + +static void _lib_put(struct dso_data *data) +{ + if (!--data->ref_count) { + dlclose(data->dso_handle); + UNLINK_DSO(data); + _free_dso_data(data); + } +} + +/* Find DSO data. */ +static struct dso_data *_lookup_dso(struct message_data *data) +{ + struct dso_data *dso_data, *ret = NULL; + + dm_list_iterate_items(dso_data, &_dso_registry) + if (!strcmp(data->dso_name, dso_data->dso_name)) { + _lib_get(dso_data); + ret = dso_data; + break; + } + + return ret; +} + +/* Lookup DSO symbols we need. */ +static int _lookup_symbol(void *dl, void **symbol, const char *name) +{ + if (!(*symbol = dlsym(dl, name))) + return_0; + + return 1; +} + +static int _lookup_symbols(void *dl, struct dso_data *data) +{ + return _lookup_symbol(dl, (void *) &data->process_event, + "process_event") && + _lookup_symbol(dl, (void *) &data->register_device, + "register_device") && + _lookup_symbol(dl, (void *) &data->unregister_device, + "unregister_device"); +} + +/* Load an application specific DSO. */ +static struct dso_data *_load_dso(struct message_data *data) +{ + void *dl; + struct dso_data *ret; + const char *dlerr; + + if (!(dl = dlopen(data->dso_name, RTLD_NOW))) { + dlerr = dlerror(); + goto_bad; + } + + if (!(ret = _alloc_dso_data(data))) { + dlclose(dl); + dlerr = "no memory"; + goto_bad; + } + + if (!(_lookup_symbols(dl, ret))) { + _free_dso_data(ret); + dlclose(dl); + dlerr = "symbols missing"; + goto_bad; + } + + /* + * Keep handle to close the library once + * we've got no references to it any more. + */ + ret->dso_handle = dl; + LINK_DSO(ret); return ret; +bad: + log_error("dmeventd %s dlopen failed: %s.", data->dso_name, dlerr); + data->msg->size = dm_asprintf(&(data->msg->data), "%s %s dlopen failed: %s", + data->id, data->dso_name, dlerr); + return NULL; } -static void _lib_put(struct dso_data *data); +/************ + * THREAD + ************/ + +/* Allocate/free the thread status structure for a monitoring thread. */ static void _free_thread_status(struct thread_status *thread) { _lib_put(thread->dso_data); @@ -276,19 +374,25 @@ static void _free_thread_status(struct thread_status *thread) dm_free(thread); } -/* Allocate/free DSO data. */ -static struct dso_data *_alloc_dso_data(struct message_data *data) +/* Allocate/free the status structure for a monitoring thread. */ +static struct thread_status *_alloc_thread_status(const struct message_data *data, + struct dso_data *dso_data) { - struct dso_data *ret = (typeof(ret)) dm_zalloc(sizeof(*ret)); + struct thread_status *ret; - if (!ret) + if (!(ret = dm_zalloc(sizeof(*ret)))) return NULL; - if (!(ret->dso_name = dm_strdup(data->dso_name))) { + if (!(ret->device.uuid = dm_strdup(data->device_uuid))) { dm_free(ret); return NULL; } + ret->dso_data = dso_data; + ret->events = data->events_field; + ret->timeout = data->timeout_secs; + dm_list_init(&ret->timeout_list); + return ret; } @@ -331,12 +435,6 @@ static int _pthread_create_smallstack(pthread_t *t, void *(*fun)(void *), void * return r; } -static void _free_dso_data(struct dso_data *data) -{ - dm_free(data->dso_name); - dm_free(data); -} - /* * Fetch a string off src and duplicate it into *ptr. * Pay attention to zero-length strings. @@ -893,96 +991,6 @@ static int _terminate_thread(struct thread_status *thread) return pthread_kill(thread->thread, SIGALRM); } -/* DSO reference counting. Call with _global_mutex locked! */ -static void _lib_get(struct dso_data *data) -{ - data->ref_count++; -} - -static void _lib_put(struct dso_data *data) -{ - if (!--data->ref_count) { - dlclose(data->dso_handle); - UNLINK_DSO(data); - _free_dso_data(data); - } -} - -/* Find DSO data. */ -static struct dso_data *_lookup_dso(struct message_data *data) -{ - struct dso_data *dso_data, *ret = NULL; - - dm_list_iterate_items(dso_data, &_dso_registry) - if (!strcmp(data->dso_name, dso_data->dso_name)) { - _lib_get(dso_data); - ret = dso_data; - break; - } - - return ret; -} - -/* Lookup DSO symbols we need. */ -static int _lookup_symbol(void *dl, void **symbol, const char *name) -{ - if ((*symbol = dlsym(dl, name))) - return 1; - - return 0; -} - -static int _lookup_symbols(void *dl, struct dso_data *data) -{ - return _lookup_symbol(dl, (void *) &data->process_event, - "process_event") && - _lookup_symbol(dl, (void *) &data->register_device, - "register_device") && - _lookup_symbol(dl, (void *) &data->unregister_device, - "unregister_device"); -} - -/* Load an application specific DSO. */ -static struct dso_data *_load_dso(struct message_data *data) -{ - void *dl; - struct dso_data *ret; - - if (!(dl = dlopen(data->dso_name, RTLD_NOW))) { - const char *dlerr = dlerror(); - log_error("dmeventd %s dlopen failed: %s.", - data->dso_name, dlerr); - data->msg->size = - dm_asprintf(&(data->msg->data), "%s %s dlopen failed: %s", - data->id, data->dso_name, dlerr); - return NULL; - } - - if (!(ret = _alloc_dso_data(data))) { - dlclose(dl); - return NULL; - } - - if (!(_lookup_symbols(dl, ret))) { - _free_dso_data(ret); - dlclose(dl); - return NULL; - } - - /* - * Keep handle to close the library once - * we've got no references to it any more. - */ - ret->dso_handle = dl; - _lib_get(ret); - - _lock_mutex(); - LINK_DSO(ret); - _unlock_mutex(); - - return ret; -} - /* Return success on daemon active check. */ static int _active(struct message_data *message_data) {

1 0

master - dmeventd: using warning level
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=590091a4faa4a4eb… Commit: 590091a4faa4a4eb66598137576e01e692fb9f8d Parent: 9488cbdd0bd3a410ec1e77454c6a18bdfe8612f5 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 11:14:12 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:33:19 2015 +0200 dmeventd: using warning level When dmevend notices problems, but continues to operate normally change log level to warning. --- daemons/dmeventd/dmeventd.c | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index f2c417c..93df32e 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -2155,7 +2155,7 @@ int main(int argc, char *argv[]) #ifdef __linux__ /* Systemd has adjusted oom killer for us already */ if (!_systemd_activation && !_protect_against_oom_killer()) - log_error("Failed to protect against OOM killer."); + log_warn("WARNING: Failed to protect against OOM killer."); #endif _init_thread_signals(); @@ -2187,8 +2187,8 @@ int main(int argc, char *argv[]) _unlock_mutex(); if (nothreads) break; - log_error("There are still devices being monitored."); - log_error("Refusing to exit."); + log_warn("WARNING: There are still devices being monitored."); + log_warn("WARNING: Refusing to exit."); } _process_request(&fifos); _cleanup_unused_threads();

1 0

master - dmeventd: no registering of 0 event mask
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=9488cbdd0bd3a410… Commit: 9488cbdd0bd3a410ec1e77454c6a18bdfe8612f5 Parent: fa9e41d2e373e7e8017b158f60eb3e77a582f532 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 11:10:57 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:33:19 2015 +0200 dmeventd: no registering of 0 event mask Whenever user tries to register 0 mask report this as EINVAL. --- daemons/dmeventd/dmeventd.c | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 80e6b14..f2c417c 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -1472,6 +1472,8 @@ static int _handle_request(struct dm_event_daemon_message *msg, { switch (msg->cmd) { case DM_EVENT_CMD_REGISTER_FOR_EVENT: + if (!message_data->events_field) + return -EINVAL; return _register_for_event(message_data); case DM_EVENT_CMD_UNREGISTER_FOR_EVENT: return _unregister_for_event(message_data);

1 0

master - dmeventd: thin plugin update
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=fa9e41d2e373e7e8… Commit: fa9e41d2e373e7e8017b158f60eb3e77a582f532 Parent: 6b0bc5b2d9cc2f160b14d34e764bff99cd3c595f Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 20 14:19:35 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:33:07 2015 +0200 dmeventd: thin plugin update Use dm_make_percent for percentage calculation like lvm2 command. Use a single call for resize. --- WHATS_NEW_DM | 1 + daemons/dmeventd/plugins/thin/dmeventd_thin.c | 86 ++++++++++++------------ 2 files changed, 44 insertions(+), 43 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index ba22f1e..2b87e4b 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Thin plugin for dmeventd improved percentage usage. Snapshot plugin for dmeventd improved percentage usage. Add dm_hold_control_dev to allow holding of control device open. Add dm_report_compact_given_fields to remove given empty fields from report. diff --git a/daemons/dmeventd/plugins/thin/dmeventd_thin.c b/daemons/dmeventd/plugins/thin/dmeventd_thin.c index 4f08d5b..888cc0c 100644 --- a/daemons/dmeventd/plugins/thin/dmeventd_thin.c +++ b/daemons/dmeventd/plugins/thin/dmeventd_thin.c @@ -28,11 +28,11 @@ #endif /* First warning when thin data or metadata is 80% full. */ -#define WARNING_THRESH 80 +#define WARNING_THRESH (DM_PERCENT_1 * 80) /* Run a check every 5%. */ -#define CHECK_STEP 5 +#define CHECK_STEP (DM_PERCENT_1 * 5) /* Do not bother checking thin data or metadata is less than 50% full. */ -#define CHECK_MINIMUM 50 +#define CHECK_MINIMUM (DM_PERCENT_1 * 50) #define UMOUNT_COMMAND "/bin/umount" @@ -138,14 +138,6 @@ out: return r; } -static int _extend(struct dso_state *state) -{ -#if THIN_DEBUG - log_info("dmeventd executes: %s.", state->cmd_str); -#endif - return dmeventd_lvm2_run_with_lock(state->cmd_str); -} - static int _run(const char *cmd, ...) { va_list ap; @@ -188,9 +180,9 @@ static int _run(const char *cmd, ...) } struct mountinfo_s { + const char *device; struct dm_info info; dm_bitset_t minors; /* Bitset for active thin pool minors */ - const char *device; }; static int _umount_device(char *buffer, unsigned major, unsigned minor, @@ -213,24 +205,24 @@ static int _umount_device(char *buffer, unsigned major, unsigned minor, * Find all thin pool users and try to umount them. * TODO: work with read-only thin pool support */ -static void _umount(struct dm_task *dmt, const char *device) +static void _umount(struct dm_task *dmt) { /* TODO: Convert to use hash to reduce memory usage */ static const size_t MINORS = (1U << 20); /* 20 bit */ - struct mountinfo_s data = { - .device = device, - }; + struct mountinfo_s data = { NULL }; if (!dm_task_get_info(dmt, &data.info)) return; + data.device = dm_task_get_name(dmt); + if (!(data.minors = dm_bitset_create(NULL, MINORS))) { - log_error("Failed to allocate bitset. Not unmounting %s.", device); + log_error("Failed to allocate bitset. Not unmounting %s.", data.device); goto out; } if (!_find_all_devs(data.minors, data.info.major, data.info.minor)) { - log_error("Failed to detect mounted volumes for %s.", device); + log_error("Failed to detect mounted volumes for %s.", data.device); goto out; } @@ -244,6 +236,18 @@ out: dm_bitset_destroy(data.minors); } +static void _use_policy(struct dm_task *dmt, struct dso_state *state) +{ +#if THIN_DEBUG + log_info("dmeventd executes: %s.", state->cmd_str); +#endif + if (!dmeventd_lvm2_run_with_lock(state->cmd_str)) { + log_error("Failed to extend thin pool %s.", + dm_task_get_name(dmt)); + _umount(dmt); + } +} + void process_event(struct dm_task *dmt, enum dm_event_mask event __attribute__((unused)), void **user) @@ -256,12 +260,19 @@ void process_event(struct dm_task *dmt, uint64_t start, length; char *target_type = NULL; char *params; + int needs_policy = 0; #if 0 /* No longer monitoring, waiting for remove */ if (!state->meta_percent_check && !state->data_percent_check) return; #endif + if (event & DM_EVENT_DEVICE_ERROR) { + /* Error -> no need to check and do instant resize */ + _use_policy(dmt, state); + return; + } + dm_get_next_target(dmt, next, &start, &length, &target_type, &params); if (!target_type || (strcmp(target_type, "thin-pool") != 0)) { @@ -271,13 +282,13 @@ void process_event(struct dm_task *dmt, if (!dm_get_status_thin_pool(state->mem, params, &tps)) { log_error("Failed to parse status."); - _umount(dmt, device); + _umount(dmt); goto out; } #if THIN_DEBUG - log_debug("%p: Got status " FMTu64 " / " FMTu64 " " FMTu64 - " / " FMTu64 ".", state, + log_debug("Thin pool status " FMTu64 "/" FMTu64 " " + FMTu64 "/" FMTu64 ".", tps->used_metadata_blocks, tps->total_metadata_blocks, tps->used_data_blocks, tps->total_data_blocks); #endif @@ -293,7 +304,7 @@ void process_event(struct dm_task *dmt, state->known_data_size = tps->total_data_blocks; } - percent = 100 * tps->used_metadata_blocks / tps->total_metadata_blocks; + percent = dm_make_percent(tps->used_metadata_blocks, tps->total_metadata_blocks); if (percent >= state->metadata_percent_check) { /* * Usage has raised more than CHECK_STEP since the last @@ -303,18 +314,12 @@ void process_event(struct dm_task *dmt, /* FIXME: extension of metadata needs to be written! */ if (percent >= WARNING_THRESH) /* Print a warning to syslog. */ - log_warn("WARNING: Thin metadata %s is now %i%% full.", - device, percent); - /* Try to extend the metadata, in accord with user-set policies */ - if (!_extend(state)) { - log_error("Failed to extend thin metadata %s.", - device); - _umount(dmt, device); - } - /* FIXME: hmm READ-ONLY switch should happen in error path */ + log_warn("WARNING: Thin pool %s metadata is now %.2f%% full.", + device, dm_percent_to_float(percent)); + needs_policy = 1; } - percent = 100 * tps->used_data_blocks / tps->total_data_blocks; + percent = dm_make_percent(tps->used_data_blocks, tps->total_data_blocks); if (percent >= state->data_percent_check) { /* * Usage has raised more than CHECK_STEP since @@ -323,21 +328,16 @@ void process_event(struct dm_task *dmt, state->data_percent_check = (percent / CHECK_STEP) * CHECK_STEP + CHECK_STEP; if (percent >= WARNING_THRESH) /* Print a warning to syslog. */ - log_warn("WARNING: Thin %s is now %i%% full.", - device, percent); - /* Try to extend the thin data, in accord with user-set policies */ - if (!_extend(state)) { - log_error("Failed to extend thin %s.", device); - state->data_percent_check = 0; - _umount(dmt, device); - } - /* FIXME: hmm READ-ONLY switch should happen in error path */ + log_warn("WARNING: Thin pool %s data is now %.2f%% full.", + device, dm_percent_to_float(percent)); + needs_policy = 1; } + + if (needs_policy) + _use_policy(dmt, state); out: if (tps) dm_pool_free(state->mem, tps); - - dmeventd_lvm2_unlock(); } int register_device(const char *device,

1 0

master - dmeventd: snapshot plugin device removal
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=6b0bc5b2d9cc2f16… Commit: 6b0bc5b2d9cc2f160b14d34e764bff99cd3c595f Parent: 7ff5b03e5e82b31ed332ee95631cede836224621 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 15:38:24 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:29:53 2015 +0200 dmeventd: snapshot plugin device removal Add #ifdef-ed code to have ability to even remove unusable device. For now purely experimental. --- .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 39 ++++++++++++++++++++ 1 files changed, 39 insertions(+), 0 deletions(-) diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index f864bb4..d9bef5c 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -84,6 +84,41 @@ static int _extend(const char *cmd) return dmeventd_lvm2_run_with_lock(cmd); } +#ifdef SNAPSHOT_REMOVE +/* Remove invalid snapshot from dm-table */ +/* Experimental for now and not used by default */ +static int _remove(const char *uuid) +{ + int r = 1; + uint32_t cookie = 0; + struct dm_task *dmt; + + if (!(dmt = dm_task_create(DM_DEVICE_REMOVE))) + return 0; + + if (!dm_task_set_uuid(dmt, uuid)) { + r = 0; + goto_out; + } + + dm_task_retry_remove(dmt); + + if (!dm_task_set_cookie(dmt, &cookie, 0)) { + r = 0; + goto_out; + } + + if (!dm_task_run(dmt)) { + r = 0; + goto_out; + } +out: + dm_task_destroy(dmt); + + return r; +} +#endif /* SNAPSHOT_REMOVE */ + static void _umount(const char *device, int major, int minor) { FILE *mounts; @@ -164,6 +199,10 @@ void process_event(struct dm_task *dmt, state->percent_check = 0; if (dm_task_get_info(dmt, &info)) _umount(device, info.major, info.minor); +#ifdef SNAPSHOT_REMOVE + /* Maybe configurable ? */ + _remove(dm_task_get_uuid(dmt)); +#endif goto out; }

1 0

master - dmeventd: snapshot plugin updates
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=7ff5b03e5e82b31e… Commit: 7ff5b03e5e82b31ed332ee95631cede836224621 Parent: 91350f5c6ad149a565402254688c6bd1700c5bf5 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 15:35:55 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:29:50 2015 +0200 dmeventd: snapshot plugin updates Improve test for invalid snapshot. Use dm_make_percent() to manipulate with exactly same percentage as lvm2 command is using. --- WHATS_NEW_DM | 1 + .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 42 +++++++++---------- 2 files changed, 21 insertions(+), 22 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 653e32d..ba22f1e 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Snapshot plugin for dmeventd improved percentage usage. Add dm_hold_control_dev to allow holding of control device open. Add dm_report_compact_given_fields to remove given empty fields from report. Use libdm status parsing and local mem raid dmeventd plugin. diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index d5575f3..f864bb4 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -20,17 +20,17 @@ #include <stdarg.h> /* First warning when snapshot is 80% full. */ -#define WARNING_THRESH 80 +#define WARNING_THRESH (DM_PERCENT_1 * 80) /* Run a check every 5%. */ -#define CHECK_STEP 5 +#define CHECK_STEP (DM_PERCENT_1 * 5) /* Do not bother checking snapshots less than 50% full. */ -#define CHECK_MINIMUM 50 +#define CHECK_MINIMUM (DM_PERCENT_1 * 50) #define UMOUNT_COMMAND "/bin/umount" struct dso_state { struct dm_pool *mem; - int percent_check; + dm_percent_t percent_check; uint64_t known_size; char cmd_lvextend[512]; }; @@ -137,6 +137,7 @@ void process_event(struct dm_task *dmt, struct dm_status_snapshot *status = NULL; const char *device = dm_task_get_name(dmt); int percent; + struct dm_info info; /* No longer monitoring, waiting for remove */ if (!state->percent_check) @@ -148,22 +149,9 @@ void process_event(struct dm_task *dmt, return; } - if (!dm_get_status_snapshot(state->mem, params, &status)) + if (!dm_get_status_snapshot(state->mem, params, &status)) { + log_error("Cannot parse snapshot %s state: %s.", device, params); return; - - if (status->invalid || status->overflow) { - struct dm_info info; - log_error("Snapshot %s is lost.", device); - if (dm_task_get_info(dmt, &info)) { - _umount(device, info.major, info.minor); - goto_out; - } /* else; too bad, but this is best-effort thing... */ - } - - /* Snapshot size had changed. Clear the threshold. */ - if (state->known_size != status->total_sectors) { - state->percent_check = CHECK_MINIMUM; - state->known_size = status->total_sectors; } /* @@ -171,19 +159,29 @@ void process_event(struct dm_task *dmt, * the status string. Report the full status string to syslog. */ if (status->invalid || status->overflow || !status->total_sectors) { - log_error("Snapshot %s changed state to: %s.", device, params); + log_warn("WARNING: Snapshot %s changed state to: %s and should be removed.", + device, params); state->percent_check = 0; + if (dm_task_get_info(dmt, &info)) + _umount(device, info.major, info.minor); goto out; } - percent = (int) (100 * status->used_sectors / status->total_sectors); + /* Snapshot size had changed. Clear the threshold. */ + if (state->known_size != status->total_sectors) { + state->percent_check = CHECK_MINIMUM; + state->known_size = status->total_sectors; + } + + percent = dm_make_percent(status->used_sectors, status->total_sectors); if (percent >= state->percent_check) { /* Usage has raised more than CHECK_STEP since the last time. Run actions. */ state->percent_check = (percent / CHECK_STEP) * CHECK_STEP + CHECK_STEP; if (percent >= WARNING_THRESH) /* Print a warning to syslog. */ - log_warn("WARNING: Snapshot %s is now %i%% full.", device, percent); + log_warn("WARNING: Snapshot %s is now %.2f%% full.", + device, dm_percent_to_float(percent)); /* Try to extend the snapshot, in accord with user-set policies */ if (!_extend(state->cmd_lvextend))

1 0

master - dmeventd: mirror plugin update
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=91350f5c6ad149a5… Commit: 91350f5c6ad149a565402254688c6bd1700c5bf5 Parent: 9c5c9e2355826ad3835f35e494dde9bb8b1e6356 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 17:09:43 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:28:37 2015 +0200 dmeventd: mirror plugin update Don't use --config - this requires reload of lvm.conf --- daemons/dmeventd/plugins/mirror/dmeventd_mirror.c | 3 +-- 1 files changed, 1 insertions(+), 2 deletions(-) diff --git a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c index 62ad331..e20b249 100644 --- a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c +++ b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c @@ -220,8 +220,7 @@ int register_device(const char *device, } if (!dmeventd_lvm2_command(state->mem, state->cmd_lvconvert, sizeof(state->cmd_lvconvert), - "lvconvert --config devices{ignore_suspended_devices=1} " - "--repair --use-policies", device)) { + "lvconvert --repair --use-policies", device)) { dmeventd_lvm2_exit_with_pool(state); goto_bad; }

1 0

master - dmeventd: raid plugin reporting
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=9c5c9e2355826ad3… Commit: 9c5c9e2355826ad3835f35e494dde9bb8b1e6356 Parent: cde12cbe9e4d3cabefc8644846e86b73b6217b6d Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 10:38:40 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:28:37 2015 +0200 dmeventd: raid plugin reporting Fix raid logging introduced with last updating commit. --- daemons/dmeventd/plugins/raid/dmeventd_raid.c | 9 ++++++--- 1 files changed, 6 insertions(+), 3 deletions(-) diff --git a/daemons/dmeventd/plugins/raid/dmeventd_raid.c b/daemons/dmeventd/plugins/raid/dmeventd_raid.c index f00ffa4..2f200ba 100644 --- a/daemons/dmeventd/plugins/raid/dmeventd_raid.c +++ b/daemons/dmeventd/plugins/raid/dmeventd_raid.c @@ -37,14 +37,17 @@ static int _process_raid_event(struct dso_state *state, char *params, const char return 0; } - if ((d = strchr(status->dev_health, 'D')) && !state->failed) { + if ((d = strchr(status->dev_health, 'D'))) { + if (state->failed) + goto out; /* already reported */ + log_error("Device #%d of %s array, %s, has failed.", (int)(d - status->dev_health), status->raid_type, device); state->failed = 1; if (!dmeventd_lvm2_run_with_lock(state->cmd_lvscan)) - log_info("Re-scan of RAID device %s failed.", device); + log_warn("WARNING: Re-scan of RAID device %s failed.", device); /* if repair goes OK, report success even if lvscan has failed */ if (!dmeventd_lvm2_run_with_lock(state->cmd_lvconvert)) { @@ -58,7 +61,7 @@ static int _process_raid_event(struct dso_state *state, char *params, const char status->raid_type, device, (status->insync_regions == status->total_regions) ? "now" : "not"); } - +out: dm_pool_free(state->mem, status); return 1;

1 0

master - dmeventd: lvm2 plugin correctly debug
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=cde12cbe9e4d3cab… Commit: cde12cbe9e4d3cabefc8644846e86b73b6217b6d Parent: ab6d16a8a5214ace81b9af346b9c065ec62fb04b Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 10:38:05 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:28:37 2015 +0200 dmeventd: lvm2 plugin correctly debug Fix debug message and report exit when really doing it. Also add missing '_' to static function. --- daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c index 49bfb56..aaf9a49 100644 --- a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c +++ b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c @@ -34,8 +34,8 @@ static void *_lvm_handle = NULL; DM_EVENT_LOG_FN("lvm") -static void lvm2_print_log(int level, const char *file, int line, - int dm_errno_or_class, const char *msg) +static void _lvm2_print_log(int level, const char *file, int line, + int dm_errno_or_class, const char *msg) { print_log(level, file, line, dm_errno_or_class, "%s", msg); } @@ -62,7 +62,7 @@ int dmeventd_lvm2_init(void) pthread_mutex_lock(&_register_mutex); if (!_lvm_handle) { - lvm2_log_fn(lvm2_print_log); + lvm2_log_fn(_lvm2_print_log); if (!(_lvm_handle = lvm2_init())) goto out; @@ -103,10 +103,10 @@ void dmeventd_lvm2_exit(void) log_debug("lvm plugin exiting."); lvm2_exit(_lvm_handle); _lvm_handle = NULL; + log_debug("lvm plugin exited."); } pthread_mutex_unlock(&_register_mutex); - log_debug("lvm plugin exited."); } struct dm_pool *dmeventd_lvm2_pool(void)

1 0

master - dmeventd: check for malloc return
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=ab6d16a8a5214ace… Commit: ab6d16a8a5214ace81b9af346b9c065ec62fb04b Parent: 09a62cca0ce259b207e85683dae47ac145590825 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 19 10:59:09 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:28:37 2015 +0200 dmeventd: check for malloc return --- daemons/dmeventd/dmeventd.c | 9 ++++++--- 1 files changed, 6 insertions(+), 3 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 56c504a..80e6b14 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -1394,10 +1394,13 @@ static int _client_read(struct dm_event_fifos *fifos, bytes += ret > 0 ? ret : 0; if (header && (bytes == 2 * sizeof(uint32_t))) { msg->cmd = ntohl(header[0]); - msg->size = ntohl(header[1]); - buf = msg->data = dm_malloc(msg->size); - size = msg->size; + size = msg->size = ntohl(header[1]); bytes = 0; + if (!size) + break; /* No data -> error */ + buf = msg->data = dm_malloc(msg->size); + if (!buf) + break; /* No mem -> error */ header = 0; } }

1 0

master - libdm: add dm_hold_control_dev
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=09a62cca0ce259b2… Commit: 09a62cca0ce259b207e85683dae47ac145590825 Parent: 075f85dcb518d7d9c2452af4165f44ba991a1258 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Wed Oct 21 20:36:15 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:27:31 2015 +0200 libdm: add dm_hold_control_dev Support hold of control device open. Useful for daemons so the control device is not frequently reopenned. --- WHATS_NEW_DM | 1 + libdm/.exported_symbols.DM_1_02_110 | 1 + libdm/ioctl/libdm-iface.c | 12 +++++++++++- libdm/libdevmapper.h | 3 +++ 4 files changed, 16 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 9818846..653e32d 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Add dm_hold_control_dev to allow holding of control device open. Add dm_report_compact_given_fields to remove given empty fields from report. Use libdm status parsing and local mem raid dmeventd plugin. Use local mem pool and lock only lvm2 execution for mirror dmeventd plugin. diff --git a/libdm/.exported_symbols.DM_1_02_110 b/libdm/.exported_symbols.DM_1_02_110 index 71d04a9..da742ee 100644 --- a/libdm/.exported_symbols.DM_1_02_110 +++ b/libdm/.exported_symbols.DM_1_02_110 @@ -1 +1,2 @@ dm_report_compact_given_fields +dm_hold_control_dev diff --git a/libdm/ioctl/libdm-iface.c b/libdm/ioctl/libdm-iface.c index e9566c2..58a8dfe 100644 --- a/libdm/ioctl/libdm-iface.c +++ b/libdm/ioctl/libdm-iface.c @@ -81,6 +81,7 @@ static dm_bitset_t _dm_bitset = NULL; static uint32_t _dm_device_major = 0; static int _control_fd = -1; +static int _hold_control_fd_open = 0; static int _version_checked = 0; static int _version_ok = 1; static unsigned _ioctl_buffer_double_factor = 0; @@ -2079,9 +2080,18 @@ repeat_ioctl: return 0; } +void dm_hold_control_dev(int hold_open) +{ + _hold_control_fd_open = hold_open ? 1 : 0; + + log_debug("Hold of control device is now %sset.", + _hold_control_fd_open ? "" : "un"); +} + void dm_lib_release(void) { - _close_control_fd(); + if (!_hold_control_fd_open) + _close_control_fd(); dm_timestamp_destroy(_dm_ioctl_timestamp); _dm_ioctl_timestamp = NULL; update_devs(); diff --git a/libdm/libdevmapper.h b/libdm/libdevmapper.h index fcbf7cf..08b24fe 100644 --- a/libdm/libdevmapper.h +++ b/libdm/libdevmapper.h @@ -1168,6 +1168,9 @@ void dm_lib_init(void) __attribute__((constructor)); void dm_lib_release(void); void dm_lib_exit(void) __attribute__((destructor)); +/* An optimisation for clients making repeated calls involving dm ioctls */ +void dm_hold_control_dev(int hold_open); + /* * Use NULL for all devices. */

1 0

master - tools: do not change signals for memlocked daemons
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=075f85dcb518d7d9… Commit: 075f85dcb518d7d9c2452af4165f44ba991a1258 Parent: d2c4ce254bac0f607676b10e38ba2f2ebee709b8 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 10:45:58 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:26:30 2015 +0200 tools: do not change signals for memlocked daemons --- lib/misc/lvm-signal.c | 12 ++++++++++++ 1 files changed, 12 insertions(+), 0 deletions(-) diff --git a/lib/misc/lvm-signal.c b/lib/misc/lvm-signal.c index 0d61e73..9250560 100644 --- a/lib/misc/lvm-signal.c +++ b/lib/misc/lvm-signal.c @@ -15,6 +15,7 @@ #include "lib.h" #include "lvm-signal.h" +#include "memlock.h" #include <signal.h> @@ -60,6 +61,8 @@ void sigint_allow(void) struct sigaction handler; sigset_t sigs; + if (memlock_count_daemon()) + return; /* * Do not overwrite the backed-up handler data - * just increase nesting count. @@ -91,6 +94,9 @@ void sigint_allow(void) void sigint_restore(void) { + if (memlock_count_daemon()) + return; + if (!_handler_installed || --_handler_installed >= MAX_SIGINTS) return; @@ -112,6 +118,9 @@ void block_signals(uint32_t flags __attribute__((unused))) { sigset_t set; + if (memlock_count_daemon()) + return; + if (_signals_blocked) return; @@ -130,6 +139,9 @@ void block_signals(uint32_t flags __attribute__((unused))) void unblock_signals(void) { + if (memlock_count_daemon()) + return; + /* Don't unblock signals while any locks are held */ if (!_signals_blocked) return;

1 0

master - tools: preselect some setting for memlocked daemons
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=d2c4ce254bac0f60… Commit: d2c4ce254bac0f607676b10e38ba2f2ebee709b8 Parent: 7b78d496bf4e80dab82824a70b1d1d852e7cc1fb Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 10:44:38 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:26:30 2015 +0200 tools: preselect some setting for memlocked daemons When our daemon is locked into memory - avoid communication with dmeventd and also skip suspended devices. --- lib/misc/lvm-globals.c | 6 ++++-- tools/lvmcmdlib.c | 6 ++++-- 2 files changed, 8 insertions(+), 4 deletions(-) diff --git a/lib/misc/lvm-globals.c b/lib/misc/lvm-globals.c index 60397e5..0cdcc88 100644 --- a/lib/misc/lvm-globals.c +++ b/lib/misc/lvm-globals.c @@ -123,7 +123,8 @@ void init_mirror_in_sync(int in_sync) void init_dmeventd_monitor(int reg) { - _dmeventd_monitor = reg; + if (!memlock_count_daemon()) + _dmeventd_monitor = reg; } void init_background_polling(int polling) @@ -133,7 +134,8 @@ void init_background_polling(int polling) void init_ignore_suspended_devices(int ignore) { - _ignore_suspended_devices = ignore; + if (!memlock_count_daemon()) + _ignore_suspended_devices = ignore; } void init_ignore_lvm_mirrors(int scan) diff --git a/tools/lvmcmdlib.c b/tools/lvmcmdlib.c index 26f160f..ae8ff5c 100644 --- a/tools/lvmcmdlib.c +++ b/tools/lvmcmdlib.c @@ -79,9 +79,11 @@ int lvm2_run(void *handle, const char *cmdline) /* FIXME Temporary - move to libdevmapper */ ret = ECMD_PROCESSED; - if (!strcmp(cmdline, "_memlock_inc")) + if (!strcmp(cmdline, "_memlock_inc")) { + init_dmeventd_monitor(DMEVENTD_MONITOR_IGNORE); + init_ignore_suspended_devices(1); memlock_inc_daemon(cmd); - else if (!strcmp(cmdline, "_memlock_dec")) + } else if (!strcmp(cmdline, "_memlock_dec")) memlock_dec_daemon(cmd); else ret = lvm_run_command(cmd, argc, argv);

1 0

master - memlock: report memlock daemon counter
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=7b78d496bf4e80da… Commit: 7b78d496bf4e80dab82824a70b1d1d852e7cc1fb Parent: 19e272ba53a51a0f3103e1d5e04d06b6207f0acc Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 10:43:58 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:26:27 2015 +0200 memlock: report memlock daemon counter Add internal memlock_count_daemon(). Function can be used to recognize it's being executed from daemon restricted execution inside /lib code. --- WHATS_NEW | 1 + lib/mm/memlock.c | 5 +++++ lib/mm/memlock.h | 1 + 3 files changed, 7 insertions(+), 0 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index ece92e0..8ffbeca 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Support checking of memlock daemon counter. Allow all log levels to be used with the lvmetad -l option. Add optional shutdown when idle support for lvmetad. Fix missing in-sync progress info while lvconvert used with lvmpolld. diff --git a/lib/mm/memlock.c b/lib/mm/memlock.c index 60686ea..34a6eff 100644 --- a/lib/mm/memlock.c +++ b/lib/mm/memlock.c @@ -638,3 +638,8 @@ void memlock_unlock(struct cmd_context *cmd) } #endif + +int memlock_count_daemon() +{ + return _memlock_count_daemon; +} diff --git a/lib/mm/memlock.h b/lib/mm/memlock.h index aab9c85..3ec5ed8 100644 --- a/lib/mm/memlock.h +++ b/lib/mm/memlock.h @@ -36,6 +36,7 @@ void critical_section_dec(struct cmd_context *cmd, const char *reason); int critical_section(void); void memlock_inc_daemon(struct cmd_context *cmd); void memlock_dec_daemon(struct cmd_context *cmd); +int memlock_count_daemon(void); void memlock_init(struct cmd_context *cmd); void memlock_reset(void); void memlock_unlock(struct cmd_context *cmd);

1 0

master - lib: better reporting of threshold
by Zdenek Kabelac 22 Oct '15

22 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=19e272ba53a51a0f… Commit: 19e272ba53a51a0f3103e1d5e04d06b6207f0acc Parent: 73e679f33f411c65c9b7c7e4e501d97245f0bf5d Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 22 10:46:39 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 22 22:06:14 2015 +0200 lib: better reporting of threshold Simplify code reporting warning about incorrect thresholds. --- lib/metadata/lv_manip.c | 27 ++++++++++++++++++++------- 1 files changed, 20 insertions(+), 7 deletions(-) diff --git a/lib/metadata/lv_manip.c b/lib/metadata/lv_manip.c index 7a3d9b9..316710b 100644 --- a/lib/metadata/lv_manip.c +++ b/lib/metadata/lv_manip.c @@ -4433,23 +4433,34 @@ static int _adjust_policy_params(struct cmd_context *cmd, if (lv_is_thin_pool(lv)) { policy_threshold = find_config_tree_int(cmd, activation_thin_pool_autoextend_threshold_CFG, - lv_config_profile(lv)) * DM_PERCENT_1; + lv_config_profile(lv)); policy_amount = find_config_tree_int(cmd, activation_thin_pool_autoextend_percent_CFG, lv_config_profile(lv)); - if (!policy_amount && policy_threshold < DM_PERCENT_100) { - log_error("Can't extend thin pool %s, autoextend is set to 0%%.", - display_lvname(lv)); - return 0; + if (policy_threshold < 50) { + log_warn("WARNING: Thin pool autoextend threshold %d%% is set bellow " + "minimal supported value 50%%.", policy_threshold); + policy_threshold = 50; } } else { policy_threshold = - find_config_tree_int(cmd, activation_snapshot_autoextend_threshold_CFG, NULL) * DM_PERCENT_1; + find_config_tree_int(cmd, activation_snapshot_autoextend_threshold_CFG, NULL); policy_amount = find_config_tree_int(cmd, activation_snapshot_autoextend_percent_CFG, NULL); + if (policy_threshold < 50) { + log_warn("WARNING: Snapshot autoextend threshold %d%% is set bellow " + "minimal supported value 50%%.", policy_threshold); + policy_threshold = 50; + } } - if (policy_threshold >= DM_PERCENT_100) + if (!policy_amount && policy_threshold < 100) { + log_error("Can't extend %s with %s autoextend percent set to 0%%.", + display_lvname(lv), first_seg(lv)->segtype->name); + return 0; + } + + if (policy_threshold >= 100) return 1; /* nothing to do */ if (!lv_is_active_locally(lv)) { @@ -4458,6 +4469,8 @@ static int _adjust_policy_params(struct cmd_context *cmd, return 0; } + policy_threshold *= DM_PERCENT_1; + if (lv_is_thin_pool(lv)) { if (!lv_thin_pool_percent(lv, 1, &percent)) return_0;

1 0

master - man: document possible lvmetad -l values
by David Teigland 21 Oct '15

21 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=73e679f33f411c65… Commit: 73e679f33f411c65c9b7c7e4e501d97245f0bf5d Parent: a0d819172fde16913347e5c0984dc4d67ae6befe Author: David Teigland <teigland(a)redhat.com> AuthorDate: Wed Oct 21 13:36:57 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Wed Oct 21 13:36:57 2015 -0500 man: document possible lvmetad -l values In the man page and the lvmetad help output. --- WHATS_NEW | 1 + daemons/lvmetad/lvmetad-core.c | 4 ++-- man/lvmetad.8.in | 14 +++++--------- 3 files changed, 8 insertions(+), 11 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 983b2b4..ece92e0 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Allow all log levels to be used with the lvmetad -l option. Add optional shutdown when idle support for lvmetad. Fix missing in-sync progress info while lvconvert used with lvmpolld. Add report/compact_output_cols to lvm.conf to define report cols to compact. diff --git a/daemons/lvmetad/lvmetad-core.c b/daemons/lvmetad/lvmetad-core.c index 7ec3a3a..64d998e 100644 --- a/daemons/lvmetad/lvmetad-core.c +++ b/daemons/lvmetad/lvmetad-core.c @@ -1637,11 +1637,11 @@ static int process_timeout_arg(const char *str, unsigned *max_timeouts) static void usage(const char *prog, FILE *file) { fprintf(file, "Usage:\n" - "%s [-V] [-h] [-f] [-l {all|wire|debug}] [-s path] [-t secs]\n\n" + "%s [-V] [-h] [-f] [-l level[,level ...]] [-s path] [-t secs]\n\n" " -V Show version of lvmetad\n" " -h Show this help information\n" " -f Don't fork, run in the foreground\n" - " -l Logging message level (-l {all|wire|debug})\n" + " -l Logging message levels (all,fatal,error,warn,info,wire,debug)\n" " -p Set path to the pidfile\n" " -s Set path to the socket to listen on\n" " -t Time to wait in seconds before shutdown on idle (missing or 0 = inifinite)\n\n", prog); diff --git a/man/lvmetad.8.in b/man/lvmetad.8.in index 2acfd8b..3e895ce 100644 --- a/man/lvmetad.8.in +++ b/man/lvmetad.8.in @@ -5,7 +5,7 @@ lvmetad \(em LVM metadata cache daemon .SH SYNOPSIS .B lvmetad .RB [ \-l -.RI { all | wire | debug }] +.IR level [,level...]] .RB [ \-p .IR pidfile_path ] .RB [ \-s @@ -77,15 +77,11 @@ Don't fork, but run in the foreground. .BR \-h ", " \-? Show help information. .TP -.IR \fB\-l " {" all | wire | debug } -Select the type of log messages to generate. -Messages are logged by syslog. +.B \-l \fIlevels +Specify the levels of log messages to generate as a comma separated list. +Messages are logged by syslog. Additionally, when \-f is given they are also sent to standard error. -Since release 2.02.98, there are two classes of messages: wire and debug. -Selecting 'all' supplies both and is equivalent to a comma-separated list -\-l wire,debug. -Prior to release 2.02.98, repeating \-d from 1 to 3 times, viz. \-d, \-dd, \-ddd, -increased the detail of messages. +Possible levels are: all, fatal, error, warn, info, wire, debug. .TP .B \-p \fIpidfile_path Path to the pidfile. This overrides both the built-in default

1 0

master - libdaemon: fix typo in last log enable commit
by David Teigland 21 Oct '15

21 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=a0d819172fde1691… Commit: a0d819172fde16913347e5c0984dc4d67ae6befe Parent: 0aee04288e221de91de022d84d19424d4698a05c Author: David Teigland <teigland(a)redhat.com> AuthorDate: Wed Oct 21 11:22:22 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Wed Oct 21 11:22:22 2015 -0500 libdaemon: fix typo in last log enable commit --- libdaemon/server/daemon-log.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/libdaemon/server/daemon-log.c b/libdaemon/server/daemon-log.c index abe602b..2bebf41 100644 --- a/libdaemon/server/daemon-log.c +++ b/libdaemon/server/daemon-log.c @@ -166,7 +166,7 @@ static int _parse_one(log_state *s, int outlet, const char *type, int enable) daemon_log_enable(s, outlet, DAEMON_LOG_ERROR, enable); else if (!strcmp(type, "warn")) daemon_log_enable(s, outlet, DAEMON_LOG_WARN, enable); - else if (!strcmp(type, "warn")) + else if (!strcmp(type, "info")) daemon_log_enable(s, outlet, DAEMON_LOG_INFO, enable); else if (!strcmp(type, "wire")) daemon_log_enable(s, outlet, DAEMON_LOG_WIRE, enable);

1 0

master - libdaemon: allow all debug levels for stderr
by David Teigland 21 Oct '15

21 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=0aee04288e221de9… Commit: 0aee04288e221de91de022d84d19424d4698a05c Parent: ef4d69f456b22b4e58afb9fe5ff0b09603944f35 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Wed Oct 21 11:16:59 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Wed Oct 21 11:19:18 2015 -0500 libdaemon: allow all debug levels for stderr The parsing code for log levels was missing some levels. Also ignore an unrecognized level name rather than failing. --- libdaemon/server/daemon-log.c | 10 ++++++++-- 1 files changed, 8 insertions(+), 2 deletions(-) diff --git a/libdaemon/server/daemon-log.c b/libdaemon/server/daemon-log.c index 5ea5cdc..abe602b 100644 --- a/libdaemon/server/daemon-log.c +++ b/libdaemon/server/daemon-log.c @@ -160,12 +160,18 @@ static int _parse_one(log_state *s, int outlet, const char *type, int enable) if (!strcmp(type, "all")) for (i = 0; i < 32; ++i) daemon_log_enable(s, outlet, i, enable); + else if (!strcmp(type, "fatal")) + daemon_log_enable(s, outlet, DAEMON_LOG_FATAL, enable); + else if (!strcmp(type, "error")) + daemon_log_enable(s, outlet, DAEMON_LOG_ERROR, enable); + else if (!strcmp(type, "warn")) + daemon_log_enable(s, outlet, DAEMON_LOG_WARN, enable); + else if (!strcmp(type, "warn")) + daemon_log_enable(s, outlet, DAEMON_LOG_INFO, enable); else if (!strcmp(type, "wire")) daemon_log_enable(s, outlet, DAEMON_LOG_WIRE, enable); else if (!strcmp(type, "debug")) daemon_log_enable(s, outlet, DAEMON_LOG_DEBUG, enable); - else - return 0; return 1; }

1 0

master - tests: add lvmetad shutdown on idle test
by okozina 20 Oct '15

20 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=ef4d69f456b22b4e… Commit: ef4d69f456b22b4e58afb9fe5ff0b09603944f35 Parent: c15649b3af94e4af573ada0f30e35f7d1fee5321 Author: Ondrej Kozina <okozina(a)redhat.com> AuthorDate: Tue Oct 20 15:41:42 2015 +0200 Committer: Ondrej Kozina <okozina(a)redhat.com> CommitterDate: Tue Oct 20 15:45:28 2015 +0200 tests: add lvmetad shutdown on idle test --- test/shell/lvmetad-autoshutdown.sh | 56 ++++++++++++++++++++++++++++++++++++ 1 files changed, 56 insertions(+), 0 deletions(-) diff --git a/test/shell/lvmetad-autoshutdown.sh b/test/shell/lvmetad-autoshutdown.sh new file mode 100644 index 0000000..fdee4a6 --- /dev/null +++ b/test/shell/lvmetad-autoshutdown.sh @@ -0,0 +1,56 @@ +#!/bin/sh +# Copyright (C) 2015 Red Hat, Inc. All rights reserved. +# +# This copyrighted material is made available to anyone wishing to use, +# modify, copy, or redistribute it subject to the terms and conditions +# of the GNU General Public License v.2. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software Foundation, +# Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + +. lib/inittest + +test -e LOCAL_LVMETAD || skip +test -e LOCAL_LVMPOLLD && skip + +kill -0 $(< LOCAL_LVMETAD) || die "lvmetad is already dead" + +lvmetad_timeout=3 + +aux prepare_pvs 1 + +vgcreate $vg1 "$dev1" + +kill $(< LOCAL_LVMETAD) +aux prepare_lvmetad -t $lvmetad_timeout + +sleep $lvmetad_timeout + +# lvmetad should die after timeout, but give it some time to do so +i=0 +while kill -0 $(< LOCAL_LVMETAD) 2>/dev/null; do + test $i -ge $((lvmetad_timeout*10)) && die "lvmetad didn't shutdown with optional timeout: $lvmetad_timeout seconds" + sleep .1 + i=$((i+1)) +done + +aux prepare_lvmetad -t 0 +sleep 1 +# lvmetad must not die with -t 0 option +kill -0 $(< LOCAL_LVMETAD) || die "lvmetad died" + +kill $(< LOCAL_LVMETAD) +aux prepare_lvmetad -t $lvmetad_timeout + +sleep 1 +vgs +sleep 1 +vgs +sleep 1 +vgs + +# check that connection to lvmetad resets the timeout +kill -0 $(< LOCAL_LVMETAD) || die "lvmetad died too soon" + +vgremove -ff $vg1

1 0

master - lvmetad.8.in: update man page with timeout info
by okozina 20 Oct '15

20 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=c15649b3af94e4af… Commit: c15649b3af94e4af573ada0f30e35f7d1fee5321 Parent: f67a52677bdc95209c7adfa4575c833bed1659f1 Author: Ondrej Kozina <okozina(a)redhat.com> AuthorDate: Tue Oct 20 13:33:29 2015 +0200 Committer: Ondrej Kozina <okozina(a)redhat.com> CommitterDate: Tue Oct 20 15:45:21 2015 +0200 lvmetad.8.in: update man page with timeout info --- man/lvmetad.8.in | 6 ++++++ 1 files changed, 6 insertions(+), 0 deletions(-) diff --git a/man/lvmetad.8.in b/man/lvmetad.8.in index cdaf198..2acfd8b 100644 --- a/man/lvmetad.8.in +++ b/man/lvmetad.8.in @@ -10,6 +10,8 @@ lvmetad \(em LVM metadata cache daemon .IR pidfile_path ] .RB [ \-s .IR socket_path ] +.RB [ \-t +.IR timeout_value ] .RB [ \-f ] .RB [ \-h ] .RB [ \-V ] @@ -97,6 +99,10 @@ Path to the socket file. This overrides both the built-in default \fBLVM_LVMETAD_SOCKET\fP. To communicate successfully with lvmetad, all LVM2 processes should use the same socket path. .TP +.B \-t \fItimeout_value +The daemon may shutdown after being idle for the given time (in seconds). When the +option is omitted or the value given is zero the daemon never shutdowns on idle. +.TP .B \-V Display the version of lvmetad daemon. .SH ENVIRONMENT VARIABLES

1 0

master - lvmetad: add optional timeout option
by okozina 20 Oct '15

20 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=f67a52677bdc9520… Commit: f67a52677bdc95209c7adfa4575c833bed1659f1 Parent: 392248186e56fb216e49a2fee0caad5a506e3dd1 Author: Ondrej Kozina <okozina(a)redhat.com> AuthorDate: Tue Oct 20 13:24:50 2015 +0200 Committer: Ondrej Kozina <okozina(a)redhat.com> CommitterDate: Tue Oct 20 15:45:14 2015 +0200 lvmetad: add optional timeout option --- WHATS_NEW | 1 + daemons/lvmetad/lvmetad-core.c | 40 ++++++++++++++++++++++++++++++++++++---- 2 files changed, 37 insertions(+), 4 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 763896e..983b2b4 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Add optional shutdown when idle support for lvmetad. Fix missing in-sync progress info while lvconvert used with lvmpolld. Add report/compact_output_cols to lvm.conf to define report cols to compact. Do not change logging in lvm2 library when it's already set. diff --git a/daemons/lvmetad/lvmetad-core.c b/daemons/lvmetad/lvmetad-core.c index 89aaaf8..7ec3a3a 100644 --- a/daemons/lvmetad/lvmetad-core.c +++ b/daemons/lvmetad/lvmetad-core.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2012 Red Hat, Inc. + * Copyright (C) 2012-2015 Red Hat, Inc. * * This file is part of LVM2. * @@ -24,6 +24,7 @@ #include "lvm-version.h" #include <assert.h> +#include <errno.h> #include <pthread.h> #define LVMETAD_SOCKET DEFAULT_RUN_DIR "/lvmetad.socket" @@ -123,6 +124,7 @@ struct vg_info { #define VGFL_INVALID 0x00000001 typedef struct { + daemon_idle *idle; log_state *log; /* convenience */ const char *log_config; @@ -1592,6 +1594,9 @@ static int init(daemon_state *s) /* if (ls->initial_registrations) _process_initial_registrations(ds->initial_registrations); */ + if (ls->idle) + ls->idle->is_idle = 1; + return 1; } @@ -1614,21 +1619,39 @@ static int fini(daemon_state *s) return 1; } +static int process_timeout_arg(const char *str, unsigned *max_timeouts) +{ + char *endptr; + unsigned long l; + + errno = 0; + l = strtoul(str, &endptr, 10); + if (errno || *endptr || l >= UINT_MAX) + return 0; + + *max_timeouts = (unsigned) l; + + return 1; +} + static void usage(const char *prog, FILE *file) { fprintf(file, "Usage:\n" - "%s [-V] [-h] [-f] [-l {all|wire|debug}] [-s path]\n\n" + "%s [-V] [-h] [-f] [-l {all|wire|debug}] [-s path] [-t secs]\n\n" " -V Show version of lvmetad\n" " -h Show this help information\n" " -f Don't fork, run in the foreground\n" " -l Logging message level (-l {all|wire|debug})\n" " -p Set path to the pidfile\n" - " -s Set path to the socket to listen on\n\n", prog); + " -s Set path to the socket to listen on\n" + " -t Time to wait in seconds before shutdown on idle (missing or 0 = inifinite)\n\n", prog); } int main(int argc, char *argv[]) { signed char opt; + struct timeval timeout; + daemon_idle di = { .ptimeout = &timeout }; lvmetad_state ls = { .log_config = "" }; daemon_state s = { .daemon_fini = fini, @@ -1643,7 +1666,7 @@ int main(int argc, char *argv[]) }; // use getopt_long - while ((opt = getopt(argc, argv, "?fhVl:p:s:")) != EOF) { + while ((opt = getopt(argc, argv, "?fhVl:p:s:t:")) != EOF) { switch (opt) { case 'h': usage(argv[0], stdout); @@ -1663,6 +1686,15 @@ int main(int argc, char *argv[]) case 's': // --socket s.socket_path = optarg; break; + case 't': + if (!process_timeout_arg(optarg, &di.max_timeouts)) { + fprintf(stderr, "Invalid value of timeout parameter.\n"); + exit(EXIT_FAILURE); + } + /* 0 equals to wait indefinitely */ + if (di.max_timeouts) + s.idle = ls.idle = &di; + break; case 'V': printf("lvmetad version: " LVM_VERSION "\n"); exit(1);

1 0

master - tests: add test for missing mirror in-sync info
by okozina 20 Oct '15

20 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=392248186e56fb21… Commit: 392248186e56fb216e49a2fee0caad5a506e3dd1 Parent: 33465066c5eddaa6169814d645d543e5673bc2a6 Author: Ondrej Kozina <okozina(a)redhat.com> AuthorDate: Tue Oct 20 11:55:51 2015 +0200 Committer: Ondrej Kozina <okozina(a)redhat.com> CommitterDate: Tue Oct 20 12:19:02 2015 +0200 tests: add test for missing mirror in-sync info --- test/shell/lvconvert-mirror.sh | 9 ++++++++- 1 files changed, 8 insertions(+), 1 deletions(-) diff --git a/test/shell/lvconvert-mirror.sh b/test/shell/lvconvert-mirror.sh index 844f8e9..029112c 100644 --- a/test/shell/lvconvert-mirror.sh +++ b/test/shell/lvconvert-mirror.sh @@ -1,5 +1,5 @@ #!/bin/sh -# Copyright (C) 2010-2013 Red Hat, Inc. All rights reserved. +# Copyright (C) 2010-2015 Red Hat, Inc. All rights reserved. # # This copyrighted material is made available to anyone wishing to use, # modify, copy, or redistribute it subject to the terms and conditions @@ -313,4 +313,11 @@ lvconvert --mirrorlog disk $vg/$lv1 "$dev3:0-1" check mirror $vg $lv1 "$dev3" lvremove -ff $vg +# bz1272175: check lvconvert reports progress while waiting for mirror +# to get synced +lvcreate -l2 -n $lv1 $vg +lvconvert --type mirror -i1 -m1 $vg/$lv1 | tee out +grep -e "$vg/$lv1: Converted:" out || fail "Missing sync info in foreground mode" +lvremove -ff $vg + vgremove -ff $vg

1 0

master - polldaemon: fix missing mirror in-sync info with lvmpolld
by okozina 20 Oct '15

20 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=33465066c5eddaa6… Commit: 33465066c5eddaa6169814d645d543e5673bc2a6 Parent: a5c4c4efbd18809320f49d7ed60c58883140259c Author: Ondrej Kozina <okozina(a)redhat.com> AuthorDate: Mon Oct 19 16:56:45 2015 +0200 Committer: Ondrej Kozina <okozina(a)redhat.com> CommitterDate: Tue Oct 20 12:18:55 2015 +0200 polldaemon: fix missing mirror in-sync info with lvmpolld CONVERTING status flag is a tricky one. It's not set when converting a non-mirror LV type to the mirror type, i.e.: linear -> two leg mirror. Also the conversion itself is instant and doesn't require to be polled. When mirror reaches sync state there's no final update on VG metadata for lvmpolld to be made thereby report_progress in fact doesn't report percentage of mirror being converted but percentage of mirror being in sync. Perhaps we should reword the lvconvert output here. On the other hand CONVERTING is set while we upconvert the mirror from i.e. two leg mirror to four leg mirror. In such case the operation is required to be polled so that lvmpolld can cleanup temporary conversion log when the conversion is over. Ignore CONVERTING lv_type for the moment and match LVs only by uuids during 'mirror conversion'/'waiting for a sync to finish'. --- WHATS_NEW | 1 + tools/polldaemon.c | 7 ++++++- 2 files changed, 7 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index f973f11..763896e 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Fix missing in-sync progress info while lvconvert used with lvmpolld. Add report/compact_output_cols to lvm.conf to define report cols to compact. Do not change logging in lvm2 library when it's already set. Check for enough space in thin-pool in command before creating new thin. diff --git a/tools/polldaemon.c b/tools/polldaemon.c index cee1983..ccfaa7d 100644 --- a/tools/polldaemon.c +++ b/tools/polldaemon.c @@ -412,7 +412,12 @@ static int report_progress(struct cmd_context *cmd, struct poll_operation_id *id if (lv && id->uuid && strcmp(id->uuid, (char *)&lv->lvid)) lv = NULL; - if (lv && parms->lv_type && !(lv->status & parms->lv_type)) + + /* + * CONVERTING is set only during mirror upconversion but we may need to + * read LV's progress info even when it's not converting (linear->mirror) + */ + if (lv && (parms->lv_type ^ CONVERTING) && !(lv->status & parms->lv_type)) lv = NULL; if (!lv) {

1 0

master - conf: clarify compact output settings
by Peter Rajnoha 20 Oct '15

20 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=a5c4c4efbd188093… Commit: a5c4c4efbd18809320f49d7ed60c58883140259c Parent: 83d475626a4d9ecaa2084f349cb2cee4f0f3a719 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Tue Oct 20 11:01:09 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Tue Oct 20 11:01:09 2015 +0200 conf: clarify compact output settings --- conf/example.conf.in | 21 ++++++++++++--------- lib/config/config_settings.h | 21 ++++++++++++--------- 2 files changed, 24 insertions(+), 18 deletions(-) diff --git a/conf/example.conf.in b/conf/example.conf.in index e96f03f..b8c0407 100644 --- a/conf/example.conf.in +++ b/conf/example.conf.in @@ -1494,19 +1494,22 @@ activation { # report { # Configuration option report/compact_output. - # Do not print empty report fields. - # Fields that don't have a value set for any of the rows reported are - # skipped and not printed. Compact output is applicable only if - # report/buffered is enabled. + # Do not print empty values for all report fields. + # If enabled, all fields that don't have a value set for any of the + # rows reported are skipped and not printed. Compact output is + # applicable only if report/buffered is enabled. If you need to + # compact only specified fields, use compact_output=0 and define + # report/compact_output_cols configuration setting instead. # This configuration option has an automatic default value. # compact_output = 0 # Configuration option report/compact_output_cols. - # Do not print empty values for given report fields. - # The same as compact_output setting, but the compaction is not done - # globally for all fields in report but only for given fields. If both - # compact_output and compact_output_cols is used at the same time, - # the compact_output setting prevails. + # Do not print empty values for specified report fields. + # If defined, specified fields that don't have a value set for any + # of the rows reported are skipped and not printed. Compact output + # is applicable only if report/buffered is enabled. If you need to + # compact all fields, use compact_output=1 instead in which case + # the compact_output_cols setting is then ignored. # This configuration option has an automatic default value. # compact_output_cols = "" diff --git a/lib/config/config_settings.h b/lib/config/config_settings.h index 1a261a2..dd2b407 100644 --- a/lib/config/config_settings.h +++ b/lib/config/config_settings.h @@ -1416,17 +1416,20 @@ cfg(disk_area_size_CFG, "size", disk_area_CFG_SUBSECTION, CFG_UNSUPPORTED | CFG_ cfg(disk_area_id_CFG, "id", disk_area_CFG_SUBSECTION, CFG_UNSUPPORTED | CFG_DEFAULT_UNDEFINED, CFG_TYPE_STRING, NULL, vsn(1, 0, 0), NULL, 0, NULL, NULL) cfg(report_compact_output_CFG, "compact_output", report_CFG_SECTION, CFG_PROFILABLE | CFG_DEFAULT_COMMENTED, CFG_TYPE_BOOL, DEFAULT_REP_COMPACT_OUTPUT, vsn(2, 2, 115), NULL, 0, NULL, - "Do not print empty report fields.\n" - "Fields that don't have a value set for any of the rows reported are\n" - "skipped and not printed. Compact output is applicable only if\n" - "report/buffered is enabled.\n") + "Do not print empty values for all report fields.\n" + "If enabled, all fields that don't have a value set for any of the\n" + "rows reported are skipped and not printed. Compact output is\n" + "applicable only if report/buffered is enabled. If you need to\n" + "compact only specified fields, use compact_output=0 and define\n" + "report/compact_output_cols configuration setting instead.\n") cfg(report_compact_output_cols_CFG, "compact_output_cols", report_CFG_SECTION, CFG_ALLOW_EMPTY | CFG_PROFILABLE | CFG_DEFAULT_COMMENTED, CFG_TYPE_STRING, DEFAULT_COMPACT_OUTPUT_COLS, vsn(2, 2, 133), NULL, 0, NULL, - "Do not print empty values for given report fields.\n" - "The same as compact_output setting, but the compaction is not done\n" - "globally for all fields in report but only for given fields. If both\n" - "compact_output and compact_output_cols is used at the same time,\n" - "the compact_output setting prevails.\n") + "Do not print empty values for specified report fields.\n" + "If defined, specified fields that don't have a value set for any\n" + "of the rows reported are skipped and not printed. Compact output\n" + "is applicable only if report/buffered is enabled. If you need to\n" + "compact all fields, use compact_output=1 instead in which case\n" + "the compact_output_cols setting is then ignored.\n") cfg(report_aligned_CFG, "aligned", report_CFG_SECTION, CFG_PROFILABLE | CFG_DEFAULT_COMMENTED, CFG_TYPE_BOOL, DEFAULT_REP_ALIGNED, vsn(1, 0, 0), NULL, 0, NULL, "Align columns in report output.\n")

1 0

master - pvscan: use process_each_pv
by David Teigland 19 Oct '15

19 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=83d475626a4d9eca… Commit: 83d475626a4d9ecaa2084f349cb2cee4f0f3a719 Parent: 1ea7e2634d67a7628508a677668a00f748aee0b0 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Mon Oct 19 13:58:43 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Mon Oct 19 16:15:51 2015 -0500 pvscan: use process_each_pv The old code made two loops through the PVs: in the first loop it found the max PV and VG name lengths, and in the second loop it printed each PV using the name lengths as field widths for aligning columns. The new code uses process_each_pv() which makes one loop through the PVs. In the *first* call to pvscan_single(), the max name lengths are found by looping through the lvmcache entries which have been populated by the generic process_each code prior to calling any _single functions. Subsequent calls to pvscan_single() reuse the max lengths that were found by the first call. --- lib/cache/lvmcache.c | 24 +++++++ lib/cache/lvmcache.h | 4 + tools/pvscan.c | 178 ++++++++++++++++++++++---------------------------- 3 files changed, 107 insertions(+), 99 deletions(-) diff --git a/lib/cache/lvmcache.c b/lib/cache/lvmcache.c index c495fa7..1bc4241 100644 --- a/lib/cache/lvmcache.c +++ b/lib/cache/lvmcache.c @@ -2358,3 +2358,27 @@ int lvmcache_contains_lock_type_sanlock(struct cmd_context *cmd) return 0; } +void lvmcache_get_max_name_lengths(struct cmd_context *cmd, + unsigned *pv_max_name_len, + unsigned *vg_max_name_len) +{ + struct lvmcache_vginfo *vginfo; + struct lvmcache_info *info; + unsigned len; + + *vg_max_name_len = 0; + *pv_max_name_len = 0; + + dm_list_iterate_items(vginfo, &_vginfos) { + len = strlen(vginfo->vgname); + if (*vg_max_name_len < len) + *vg_max_name_len = len; + + dm_list_iterate_items(info, &vginfo->infos) { + len = strlen(dev_name(info->dev)); + if (*pv_max_name_len < len) + *pv_max_name_len = len; + } + } +} + diff --git a/lib/cache/lvmcache.h b/lib/cache/lvmcache.h index 76b9b10..ccf3eb4 100644 --- a/lib/cache/lvmcache.h +++ b/lib/cache/lvmcache.h @@ -190,4 +190,8 @@ void lvmcache_set_preferred_duplicates(const char *vgid); int lvmcache_contains_lock_type_sanlock(struct cmd_context *cmd); +void lvmcache_get_max_name_lengths(struct cmd_context *cmd, + unsigned *pv_max_name_len, unsigned *vg_max_name_len); + + #endif diff --git a/tools/pvscan.c b/tools/pvscan.c index 3adb91c..460b74d 100644 --- a/tools/pvscan.c +++ b/tools/pvscan.c @@ -18,56 +18,66 @@ #include "lvmetad.h" #include "lvmcache.h" -unsigned pv_max_name_len = 0; -unsigned vg_max_name_len = 0; - -static void _pvscan_display_single(struct cmd_context *cmd, - struct physical_volume *pv, - void *handle __attribute__((unused))) +struct pvscan_params { + int new_pvs_found; + int pvs_found; + uint64_t size_total; + uint64_t size_new; + unsigned pv_max_name_len; + unsigned vg_max_name_len; + unsigned pv_tmp_namelen; + char *pv_tmp_name; +}; + +static int _pvscan_display_single(struct cmd_context *cmd, + struct physical_volume *pv, + struct pvscan_params *params) { /* XXXXXX-XXXX-XXXX-XXXX-XXXX-XXXX-XXXXXX */ char uuid[40] __attribute__((aligned(8))); - const unsigned suffix = sizeof(uuid) + 10; - char pv_tmp_name[pv_max_name_len + suffix]; - unsigned pv_len = pv_max_name_len; + const unsigned suffix_len = sizeof(uuid) + 10; + unsigned pv_len; const char *pvdevname = pv_dev_name(pv); /* short listing? */ if (arg_count(cmd, short_ARG) > 0) { log_print_unless_silent("%s", pvdevname); - return; + return ECMD_PROCESSED; } - if (arg_count(cmd, verbose_ARG) > 1) { - /* FIXME As per pv_display! Drop through for now. */ - /* pv_show(pv); */ + if (!params->pv_max_name_len) { + lvmcache_get_max_name_lengths(cmd, &params->pv_max_name_len, &params->vg_max_name_len); - /* FIXME - Moved to Volume Group structure */ - /* log_print("system ID %s", pv->vg->system_id); */ + params->pv_max_name_len += 2; + params->vg_max_name_len += 2; + params->pv_tmp_namelen = params->pv_max_name_len + suffix_len; - /* log_print(" "); */ - /* return; */ + if (!(params->pv_tmp_name = dm_pool_alloc(cmd->mem, params->pv_tmp_namelen))) + return ECMD_FAILED; } + pv_len = params->pv_max_name_len; + memset(params->pv_tmp_name, 0, params->pv_tmp_namelen); + if (arg_count(cmd, uuid_ARG)) { if (!id_write_format(&pv->id, uuid, sizeof(uuid))) { stack; - return; + return ECMD_FAILED; } - if (dm_snprintf(pv_tmp_name, sizeof(pv_tmp_name), "%-*s with UUID %s", - pv_max_name_len - 2, pvdevname, uuid) < 0) { + if (dm_snprintf(params->pv_tmp_name, params->pv_tmp_namelen, "%-*s with UUID %s", + params->pv_max_name_len - 2, pvdevname, uuid) < 0) { log_error("Invalid PV name with uuid."); - return; + return ECMD_FAILED; } - pvdevname = pv_tmp_name; - pv_len += suffix; + pvdevname = params->pv_tmp_name; + pv_len += suffix_len; } if (is_orphan(pv)) log_print_unless_silent("PV %-*s %-*s %s [%s]", pv_len, pvdevname, - vg_max_name_len, " ", + params->vg_max_name_len, " ", pv->fmt ? pv->fmt->name : " ", display_size(cmd, pv_size(pv))); else if (pv_status(pv) & EXPORTED_VG) @@ -78,10 +88,36 @@ static void _pvscan_display_single(struct cmd_context *cmd, else log_print_unless_silent("PV %-*s VG %-*s %s [%s / %s free]", pv_len, pvdevname, - vg_max_name_len, pv_vg_name(pv), + params->vg_max_name_len, pv_vg_name(pv), pv->fmt ? pv->fmt->name : " ", display_size(cmd, (uint64_t) pv_pe_count(pv) * pv_pe_size(pv)), display_size(cmd, (uint64_t) (pv_pe_count(pv) - pv_pe_alloc_count(pv)) * pv_pe_size(pv))); + return ECMD_PROCESSED; +} + +static int _pvscan_single(struct cmd_context *cmd, struct volume_group *vg, + struct physical_volume *pv, struct processing_handle *handle) +{ + struct pvscan_params *params = (struct pvscan_params *)handle->custom_handle; + + if ((arg_count(cmd, exported_ARG) && !(pv_status(pv) & EXPORTED_VG)) || + (arg_count(cmd, novolumegroup_ARG) && (!is_orphan(pv)))) { + return ECMD_PROCESSED; + + } + + params->pvs_found++; + + if (is_orphan(pv)) { + params->new_pvs_found++; + params->size_new += pv_size(pv); + params->size_total += pv_size(pv); + } else { + params->size_total += (uint64_t) pv_pe_count(pv) * pv_pe_size(pv); + } + + _pvscan_display_single(cmd, pv, params); + return ECMD_PROCESSED; } #define REFRESH_BEFORE_AUTOACTIVATION_RETRIES 5 @@ -327,16 +363,9 @@ out: int pvscan(struct cmd_context *cmd, int argc, char **argv) { - int new_pvs_found = 0; - int pvs_found = 0; - - struct dm_list *pvslist; - struct pv_list *pvl; - struct physical_volume *pv; - - uint64_t size_total = 0; - uint64_t size_new = 0; - unsigned len; + struct pvscan_params params = { 0 }; + struct processing_handle *handle = NULL; + int ret; if (arg_count(cmd, cache_long_ARG)) return _pvscan_lvmetad(cmd, argc, argv); @@ -377,80 +406,31 @@ int pvscan(struct cmd_context *cmd, int argc, char **argv) if (cmd->full_filter->wipe) cmd->full_filter->wipe(cmd->full_filter); - lvmcache_destroy(cmd, 1, 0); - /* populate lvmcache */ - if (!lvmetad_vg_list_to_lvmcache(cmd)) - stack; - - log_verbose("Walking through all physical volumes"); - if (!(pvslist = get_pvs(cmd))) { - unlock_vg(cmd, VG_GLOBAL); - return_ECMD_FAILED; - } - - /* eliminate exported/new if required */ - dm_list_iterate_items(pvl, pvslist) { - pv = pvl->pv; - - if ((arg_count(cmd, exported_ARG) - && !(pv_status(pv) & EXPORTED_VG)) || - (arg_count(cmd, novolumegroup_ARG) && (!is_orphan(pv)))) { - dm_list_del(&pvl->list); - free_pv_fid(pv); - continue; - } + lvmcache_destroy(cmd, 1, 0); - /* Also check for MD use? */ -/******* - if (MAJOR(pv_create_kdev_t(pv[p]->pv_name)) != MD_MAJOR) { - log_warn - ("WARNING: physical volume \"%s\" belongs to a meta device", - pv[p]->pv_name); - } - if (MAJOR(pv[p]->pv_dev) != MD_MAJOR) - continue; -********/ - pvs_found++; - - if (is_orphan(pv)) { - new_pvs_found++; - size_new += pv_size(pv); - size_total += pv_size(pv); - } else - size_total += (uint64_t) pv_pe_count(pv) * pv_pe_size(pv); + if (!(handle = init_processing_handle(cmd))) { + log_error("Failed to initialize processing handle."); + ret = ECMD_FAILED; + goto out; } - /* find maximum pv name length */ - pv_max_name_len = vg_max_name_len = 0; - dm_list_iterate_items(pvl, pvslist) { - pv = pvl->pv; - len = strlen(pv_dev_name(pv)); - if (pv_max_name_len < len) - pv_max_name_len = len; - len = strlen(pv_vg_name(pv)); - if (vg_max_name_len < len) - vg_max_name_len = len; - } - pv_max_name_len += 2; - vg_max_name_len += 2; + handle->custom_handle = &params; - dm_list_iterate_items(pvl, pvslist) { - _pvscan_display_single(cmd, pvl->pv, NULL); - free_pv_fid(pvl->pv); - } + ret = process_each_pv(cmd, argc, argv, NULL, 0, handle, _pvscan_single); - if (!pvs_found) + if (!params.pvs_found) log_print_unless_silent("No matching physical volumes found"); else log_print_unless_silent("Total: %d [%s] / in use: %d [%s] / in no VG: %d [%s]", - pvs_found, - display_size(cmd, size_total), - pvs_found - new_pvs_found, - display_size(cmd, (size_total - size_new)), - new_pvs_found, display_size(cmd, size_new)); + params.pvs_found, + display_size(cmd, params.size_total), + params.pvs_found - params.new_pvs_found, + display_size(cmd, (params.size_total - params.size_new)), + params.new_pvs_found, display_size(cmd, params.size_new)); +out: unlock_vg(cmd, VG_GLOBAL); - return ECMD_PROCESSED; + return ret; }

1 0

master - conf: regenerate
by Peter Rajnoha 19 Oct '15

19 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=1ea7e2634d67a762… Commit: 1ea7e2634d67a7628508a677668a00f748aee0b0 Parent: 23d9b17a7b06d8e6f89b2c3e9ef9f59462436458 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Mon Oct 19 10:01:14 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Mon Oct 19 10:01:14 2015 +0200 conf: regenerate --- conf/example.conf.in | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/conf/example.conf.in b/conf/example.conf.in index dddede0..e96f03f 100644 --- a/conf/example.conf.in +++ b/conf/example.conf.in @@ -389,8 +389,8 @@ allocation { # This configuration option does not have a default value defined. # Configuration section allocation/cache_settings. - # Individual settings for policies. - # See the help for individual policies for more info. + # Settings for the cache policy. + # See documentation for individual cache policies for more info. # This configuration section has an automatic default value. # cache_settings { # }

1 0

master - cleanup: typo in comment: compact_output_fields should be compact-output_cols
by Peter Rajnoha 19 Oct '15

19 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=23d9b17a7b06d8e6… Commit: 23d9b17a7b06d8e6f89b2c3e9ef9f59462436458 Parent: c3bfe07f2ad393963af5b7138bccc622e6561c8f Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Mon Oct 19 09:59:05 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Mon Oct 19 09:59:05 2015 +0200 cleanup: typo in comment: compact_output_fields should be compact-output_cols --- conf/example.conf.in | 2 +- lib/config/config_settings.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/conf/example.conf.in b/conf/example.conf.in index e1afed6..dddede0 100644 --- a/conf/example.conf.in +++ b/conf/example.conf.in @@ -1505,7 +1505,7 @@ activation { # Do not print empty values for given report fields. # The same as compact_output setting, but the compaction is not done # globally for all fields in report but only for given fields. If both - # compact_output and compact_output_fields is used at the same time, + # compact_output and compact_output_cols is used at the same time, # the compact_output setting prevails. # This configuration option has an automatic default value. # compact_output_cols = "" diff --git a/lib/config/config_settings.h b/lib/config/config_settings.h index cefc579..1a261a2 100644 --- a/lib/config/config_settings.h +++ b/lib/config/config_settings.h @@ -1425,7 +1425,7 @@ cfg(report_compact_output_cols_CFG, "compact_output_cols", report_CFG_SECTION, C "Do not print empty values for given report fields.\n" "The same as compact_output setting, but the compaction is not done\n" "globally for all fields in report but only for given fields. If both\n" - "compact_output and compact_output_fields is used at the same time,\n" + "compact_output and compact_output_cols is used at the same time,\n" "the compact_output setting prevails.\n") cfg(report_aligned_CFG, "aligned", report_CFG_SECTION, CFG_PROFILABLE | CFG_DEFAULT_COMMENTED, CFG_TYPE_BOOL, DEFAULT_REP_ALIGNED, vsn(1, 0, 0), NULL, 0, NULL,

1 0

master - config: add report/compact_output_cols to control which columns to compact in report output
by Peter Rajnoha 16 Oct '15

16 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=c3bfe07f2ad39396… Commit: c3bfe07f2ad393963af5b7138bccc622e6561c8f Parent: 508f0f5a21af8decf32944e01ecf17c0059a9c9f Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 16 16:25:51 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 16 17:05:54 2015 +0200 config: add report/compact_output_cols to control which columns to compact in report output The new report/compact_output_cols setting has exactly the same effect as report/compact_output setting. The difference is that with the new setting it's possible to define which cols should be compacted exactly in contrast to all cols in case of report/compact_output. In case both compact_output and compact_output_cols is enabled/set, the compact_output prevails. For example: $ lvmconfig --type full report/compact_output report/compact_output_cols compact_output=0 compact_output_cols="" $ lvs vg LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert lvol0 vg -wi-a----- 4.00m --- $ lvmconfig --type full report/compact_output report/compact_output_cols compact_output=0 compact_output_cols="data_percent,metadata_percent,pool_lv,move_pv,origin" $ lvs vg LV VG Attr LSize Log Cpy%Sync Convert lvol0 vg -wi-a----- 4.00m --- $ lvmconfig --type full report/compact_output report/compact_output_cols compact_output=1 compact_output_cols="data_percent,metadata_percent,pool_lv,move_pv,origin" $ lvs vg LV VG Attr LSize lvol0 vg -wi-a----- 4.00m --- WHATS_NEW | 1 + conf/example.conf.in | 9 +++++++++ lib/config/config_settings.h | 7 +++++++ lib/config/defaults.h | 2 ++ libdm/libdevmapper.h | 2 +- tools/reporter.c | 11 ++++++++--- 6 files changed, 28 insertions(+), 4 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index c089228..f973f11 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Add report/compact_output_cols to lvm.conf to define report cols to compact. Do not change logging in lvm2 library when it's already set. Check for enough space in thin-pool in command before creating new thin. Make libblkid detect all copies of the same signature if use_blkid_wiping=1. diff --git a/conf/example.conf.in b/conf/example.conf.in index c0afcb7..e1afed6 100644 --- a/conf/example.conf.in +++ b/conf/example.conf.in @@ -1501,6 +1501,15 @@ activation { # This configuration option has an automatic default value. # compact_output = 0 + # Configuration option report/compact_output_cols. + # Do not print empty values for given report fields. + # The same as compact_output setting, but the compaction is not done + # globally for all fields in report but only for given fields. If both + # compact_output and compact_output_fields is used at the same time, + # the compact_output setting prevails. + # This configuration option has an automatic default value. + # compact_output_cols = "" + # Configuration option report/aligned. # Align columns in report output. # This configuration option has an automatic default value. diff --git a/lib/config/config_settings.h b/lib/config/config_settings.h index 9cc444f..cefc579 100644 --- a/lib/config/config_settings.h +++ b/lib/config/config_settings.h @@ -1421,6 +1421,13 @@ cfg(report_compact_output_CFG, "compact_output", report_CFG_SECTION, CFG_PROFILA "skipped and not printed. Compact output is applicable only if\n" "report/buffered is enabled.\n") +cfg(report_compact_output_cols_CFG, "compact_output_cols", report_CFG_SECTION, CFG_ALLOW_EMPTY | CFG_PROFILABLE | CFG_DEFAULT_COMMENTED, CFG_TYPE_STRING, DEFAULT_COMPACT_OUTPUT_COLS, vsn(2, 2, 133), NULL, 0, NULL, + "Do not print empty values for given report fields.\n" + "The same as compact_output setting, but the compaction is not done\n" + "globally for all fields in report but only for given fields. If both\n" + "compact_output and compact_output_fields is used at the same time,\n" + "the compact_output setting prevails.\n") + cfg(report_aligned_CFG, "aligned", report_CFG_SECTION, CFG_PROFILABLE | CFG_DEFAULT_COMMENTED, CFG_TYPE_BOOL, DEFAULT_REP_ALIGNED, vsn(1, 0, 0), NULL, 0, NULL, "Align columns in report output.\n") diff --git a/lib/config/defaults.h b/lib/config/defaults.h index cd5d121..8ab1fde 100644 --- a/lib/config/defaults.h +++ b/lib/config/defaults.h @@ -201,6 +201,8 @@ #define DEFAULT_REP_LIST_ITEM_SEPARATOR "," #define DEFAULT_TIME_FORMAT "%Y-%m-%d %T %z" +#define DEFAULT_COMPACT_OUTPUT_COLS "" + #define DEFAULT_LVS_COLS "lv_name,vg_name,lv_attr,lv_size,pool_lv,origin,data_percent,metadata_percent,move_pv,mirror_log,copy_percent,convert_lv" #define DEFAULT_VGS_COLS "vg_name,pv_count,lv_count,snap_count,vg_attr,vg_size,vg_free" #define DEFAULT_PVS_COLS "pv_name,vg_name,pv_fmt,pv_attr,pv_size,pv_free" diff --git a/libdm/libdevmapper.h b/libdm/libdevmapper.h index 1a2036b..fcbf7cf 100644 --- a/libdm/libdevmapper.h +++ b/libdm/libdevmapper.h @@ -2552,7 +2552,7 @@ int dm_report_compact_fields(struct dm_report *rh); * The same as dm_report_compact_fields, but for selected fields only. * The "fields" arg is comma separated list of field names (the same format * as used for "output_fields" arg in dm_report_init fn). -*/ + */ int dm_report_compact_given_fields(struct dm_report *rh, const char *fields); /* diff --git a/tools/reporter.c b/tools/reporter.c index 8f7a272..6c93739 100644 --- a/tools/reporter.c +++ b/tools/reporter.c @@ -618,6 +618,7 @@ static int _report(struct cmd_context *cmd, int argc, char **argv, unsigned args_are_pvs; int lv_info_needed, lv_segment_status_needed; int lock_global = 0; + const char *fields_to_compact; aligned = find_config_tree_bool(cmd, report_aligned_CFG, NULL); buffered = find_config_tree_bool(cmd, report_buffered_CFG, NULL); @@ -817,9 +818,13 @@ static int _report(struct cmd_context *cmd, int argc, char **argv, break; } - if (find_config_tree_bool(cmd, report_compact_output_CFG, NULL) && - !dm_report_compact_fields(report_handle)) - log_error("Failed to compact report output."); + if (find_config_tree_bool(cmd, report_compact_output_CFG, NULL)) { + if (!dm_report_compact_fields(report_handle)) + log_error("Failed to compact report output."); + } else if ((fields_to_compact = find_config_tree_str_allow_empty(cmd, report_compact_output_cols_CFG, NULL))) { + if (!dm_report_compact_given_fields(report_handle, fields_to_compact)) + log_error("Failed to compact given columns in report output."); + } dm_report_output(report_handle);

1 0

master - libdm: add dm_report_compact_given_fields
by Peter Rajnoha 16 Oct '15

16 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=508f0f5a21af8dec… Commit: 508f0f5a21af8decf32944e01ecf17c0059a9c9f Parent: df34fcdafd20ac195e588a06c8fc5a904fa71669 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 16 15:50:13 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 16 17:05:54 2015 +0200 libdm: add dm_report_compact_given_fields dm_report_compact_given_fields is the same as dm_report_compact_fields, but it processes only given fields, not all the fields in the report like dm_report_compact_field does. --- WHATS_NEW_DM | 1 + libdm/.exported_symbols.DM_1_02_110 | 1 + libdm/libdevmapper.h | 7 ++++ libdm/libdm-report.c | 62 +++++++++++++++++++++++++++++++++- 4 files changed, 69 insertions(+), 2 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index dffcd34..9818846 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Add dm_report_compact_given_fields to remove given empty fields from report. Use libdm status parsing and local mem raid dmeventd plugin. Use local mem pool and lock only lvm2 execution for mirror dmeventd plugin. Lock protect only lvm2 execution for snapshot and thin dmeventd plugin. diff --git a/libdm/.exported_symbols.DM_1_02_110 b/libdm/.exported_symbols.DM_1_02_110 new file mode 100644 index 0000000..71d04a9 --- /dev/null +++ b/libdm/.exported_symbols.DM_1_02_110 @@ -0,0 +1 @@ +dm_report_compact_given_fields diff --git a/libdm/libdevmapper.h b/libdm/libdevmapper.h index 8077867..1a2036b 100644 --- a/libdm/libdevmapper.h +++ b/libdm/libdevmapper.h @@ -2549,6 +2549,13 @@ int dm_report_object_is_selected(struct dm_report *rh, void *object, int do_outp int dm_report_compact_fields(struct dm_report *rh); /* + * The same as dm_report_compact_fields, but for selected fields only. + * The "fields" arg is comma separated list of field names (the same format + * as used for "output_fields" arg in dm_report_init fn). +*/ +int dm_report_compact_given_fields(struct dm_report *rh, const char *fields); + +/* * Returns 1 if there is no data waiting to be output. */ int dm_report_is_empty(struct dm_report *rh); diff --git a/libdm/libdm-report.c b/libdm/libdm-report.c index 681ad04..b2e70d5 100644 --- a/libdm/libdm-report.c +++ b/libdm/libdm-report.c @@ -81,6 +81,7 @@ struct dm_report { #define FLD_ASCENDING 0x00004000 #define FLD_DESCENDING 0x00008000 #define FLD_COMPACTED 0x00010000 +#define FLD_COMPACT_ONE 0x00020000 struct field_properties { struct dm_list list; @@ -2013,7 +2014,7 @@ out: return r; } -int dm_report_compact_fields(struct dm_report *rh) +static int _do_report_compact_fields(struct dm_report *rh, int global) { struct dm_report_field *field; struct field_properties *fp; @@ -2036,7 +2037,9 @@ int dm_report_compact_fields(struct dm_report *rh) * in next step... */ dm_list_iterate_items(fp, &rh->field_props) { - if (!(fp->flags & FLD_HIDDEN)) + if (fp->flags & FLD_HIDDEN) + continue; + if (global || (fp->flags & FLD_COMPACT_ONE)) fp->flags |= (FLD_COMPACTED | FLD_HIDDEN); } @@ -2065,6 +2068,61 @@ int dm_report_compact_fields(struct dm_report *rh) return 1; } +int dm_report_compact_fields(struct dm_report *rh) +{ + return _do_report_compact_fields(rh, 1); +} + +static int _field_to_compact_match(struct dm_report *rh, const char *field, size_t flen) +{ + struct field_properties *fp; + uint32_t f; + int implicit; + + if ((_get_field(rh, field, flen, &f, &implicit))) { + dm_list_iterate_items(fp, &rh->field_props) { + if ((fp->implicit == implicit) && (fp->field_num == f)) { + fp->flags |= FLD_COMPACT_ONE; + break; + } + } + return 1; + } + + return 0; +} + +static int _parse_fields_to_compact(struct dm_report *rh, const char *fields) +{ + const char *ws; /* Word start */ + const char *we = fields; /* Word end */ + + if (!fields) + return 1; + + while (*we) { + while (*we && *we == ',') + we++; + ws = we; + while (*we && *we != ',') + we++; + if (!_field_to_compact_match(rh, ws, (size_t) (we - ws))) { + log_error("dm_report: Unrecognized field: %.*s", (int) (we - ws), ws); + return 0; + } + } + + return 1; +} + +int dm_report_compact_given_fields(struct dm_report *rh, const char *fields) +{ + if (!_parse_fields_to_compact(rh, fields)) + return_0; + + return _do_report_compact_fields(rh, 0); +} + int dm_report_object(struct dm_report *rh, void *object) { return _do_report_object(rh, object, 1, NULL);

1 0

master - lvmlockd: use flag to avoid blocking in sanlock_acquire
by David Teigland 14 Oct '15

14 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=df34fcdafd20ac19… Commit: df34fcdafd20ac195e588a06c8fc5a904fa71669 Parent: a6d1c8ac651ecdc6fbdfa20f892ba318c2dddeda Author: David Teigland <teigland(a)redhat.com> AuthorDate: Wed Oct 14 14:36:46 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Wed Oct 14 14:39:29 2015 -0500 lvmlockd: use flag to avoid blocking in sanlock_acquire If a host failed while holding a sanlock lease, sanlock_acquire will by default block and wait for the lease to expire before returning. We want it to return with an error so we can retry instead of blocking, which allows us to process other lock operations. (Enclose this in an ifdef until the new flag appears in a sanlock release.) --- daemons/lvmlockd/lvmlockd-sanlock.c | 29 +++++++++++++++++++++++++++++ 1 files changed, 29 insertions(+), 0 deletions(-) diff --git a/daemons/lvmlockd/lvmlockd-sanlock.c b/daemons/lvmlockd/lvmlockd-sanlock.c index 1e691eb..e1a85b4 100644 --- a/daemons/lvmlockd/lvmlockd-sanlock.c +++ b/daemons/lvmlockd/lvmlockd-sanlock.c @@ -1392,6 +1392,15 @@ int lm_lock_sanlock(struct lockspace *ls, struct resource *r, int ld_mode, if (adopt) flags |= SANLK_ACQUIRE_ORPHAN_ONLY; +#ifdef SANLOCK_HAS_ACQUIRE_OWNER_NOWAIT + /* + * Don't block waiting for a failed lease to expire since it causes + * sanlock_acquire to block for a long time, which would prevent this + * thread from processing other lock requests. + */ + flags |= SANLK_ACQUIRE_OWNER_NOWAIT; +#endif + rv = sanlock_acquire(lms->sock, -1, flags, 1, &rs, NULL); if (rv == -EAGAIN) { @@ -1462,6 +1471,26 @@ int lm_lock_sanlock(struct lockspace *ls, struct resource *r, int ld_mode, return -EAGAIN; } +#ifdef SANLOCK_HAS_ACQUIRE_OWNER_NOWAIT + if (rv == SANLK_ACQUIRE_OWNED_RETRY) { + /* + * The lock is held by a failed host, and will eventually + * expire. If we retry we'll eventually acquire the lock + * (or find someone else has acquired it). The EAGAIN retry + * attempts for SH locks above would not be sufficient for + * the length of expiration time. We could add a longer + * retry time here to cover the full expiration time and block + * the activation command for that long. For now just return + * the standard error indicating that another host still owns + * the lease. FIXME: return a different error number so the + * command can print an different error indicating that the + * owner of the lease is in the process of expiring? + */ + log_debug("S %s R %s lock_san acquire mode %d rv %d", ls->name, r->name, ld_mode, rv); + *retry = 0; + return -EAGAIN; + } +#endif if (rv < 0) { log_error("S %s R %s lock_san acquire error %d", ls->name, r->name, rv);

1 0

master - dmeventd: use matching function
by Zdenek Kabelac 14 Oct '15

14 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=a6d1c8ac651ecdc6… Commit: a6d1c8ac651ecdc6fbdfa20f892ba318c2dddeda Parent: 7c36d7c90c77f8225432b451cb831711cea01883 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Wed Oct 14 14:19:47 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Wed Oct 14 14:25:27 2015 +0200 dmeventd: use matching function Respect lvm2_log_fn prototype. The idea of 'reusing' print_log with plain cast is causing very strange crashes with some older 'gcc' compilers. So just do it cleanly... --- daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c | 8 +++++++- 1 files changed, 7 insertions(+), 1 deletions(-) diff --git a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c index e3d8321..49bfb56 100644 --- a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c +++ b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c @@ -34,6 +34,12 @@ static void *_lvm_handle = NULL; DM_EVENT_LOG_FN("lvm") +static void lvm2_print_log(int level, const char *file, int line, + int dm_errno_or_class, const char *msg) +{ + print_log(level, file, line, dm_errno_or_class, "%s", msg); +} + /* * Currently only one event can be processed at a time. */ @@ -56,7 +62,7 @@ int dmeventd_lvm2_init(void) pthread_mutex_lock(&_register_mutex); if (!_lvm_handle) { - lvm2_log_fn((lvm2_log_fn_t)print_log); + lvm2_log_fn(lvm2_print_log); if (!(_lvm_handle = lvm2_init())) goto out;

1 0

master - thin: enforce local activation when creation new thin
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=7c36d7c90c77f822… Commit: 7c36d7c90c77f8225432b451cb831711cea01883 Parent: bbef4edd06097a71601b1470a713ded17deb7cba Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Wed Oct 14 01:00:35 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Wed Oct 14 01:00:35 2015 +0200 thin: enforce local activation when creation new thin As we need to check how full thin-pool is - require thin-pool is locally active. --- lib/metadata/lv_manip.c | 19 +++++++++++++------ 1 files changed, 13 insertions(+), 6 deletions(-) diff --git a/lib/metadata/lv_manip.c b/lib/metadata/lv_manip.c index 8a0bbcd..7a3d9b9 100644 --- a/lib/metadata/lv_manip.c +++ b/lib/metadata/lv_manip.c @@ -7088,16 +7088,23 @@ static struct logical_volume *_lv_create_an_lv(struct volume_group *vg, } if (seg_is_thin_volume(lp)) { + thin_pool_was_active = lv_is_active(pool_lv); if (lv_is_new_thin_pool(pool_lv)) { - thin_pool_was_active = lv_is_active(pool_lv); if (!check_new_thin_pool(pool_lv)) return_NULL; /* New pool is now inactive */ - } else if (!pool_below_threshold(first_seg(pool_lv))) { - log_error("Cannot create new thin volume, free space in " - "thin pool %s reached threshold.", - display_lvname(pool_lv)); - return NULL; + } else { + if (!activate_lv_excl_local(cmd, pool_lv)) { + log_error("Aborting. Failed to locally activate thin pool %s.", + display_lvname(pool_lv)); + return 0; + } + if (!pool_below_threshold(first_seg(pool_lv))) { + log_error("Cannot create new thin volume, free space in " + "thin pool %s reached threshold.", + display_lvname(pool_lv)); + return NULL; + } } }

1 0

master - makefiles: switch to rpath-link
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=bbef4edd06097a71… Commit: bbef4edd06097a71601b1470a713ded17deb7cba Parent: 3f1a3b7090d421d46f245454bc5c8ed299a9cc1d Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Wed Oct 14 00:51:55 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Wed Oct 14 00:51:55 2015 +0200 makefiles: switch to rpath-link Plain rpath is to invasive and gets into binary. We only want to provide hint for linker. --- scripts/Makefile.in | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/scripts/Makefile.in b/scripts/Makefile.in index bfa0fad..af2b299 100644 --- a/scripts/Makefile.in +++ b/scripts/Makefile.in @@ -1,5 +1,5 @@ # -# Copyright (C) 2006-2011 Red Hat, Inc. All rights reserved. +# Copyright (C) 2006-2015 Red Hat, Inc. All rights reserved. # # This file is part of LVM2. # @@ -26,7 +26,7 @@ ifeq ("@APPLIB@", "yes") DEPLIBS += $(top_builddir)/liblvm/liblvm2app.so $(top_builddir)/libdm/libdevmapper.so LDFLAGS += -L$(top_builddir)/liblvm ifeq ("@DMEVENTD@", "yes") - LDFLAGS += -Wl,-rpath,$(top_builddir)/daemons/dmeventd + LDFLAGS += -Wl,-rpath-link,$(top_builddir)/daemons/dmeventd endif LVMLIBS = @LVM2APP_LIB@ -ldevmapper endif

1 0

master - dmeventd: fix missing '!'
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=3f1a3b7090d421d4… Commit: 3f1a3b7090d421d46f245454bc5c8ed299a9cc1d Parent: a91fbe9d27a79b4be7fad72fc7a1ba2a976ecd41 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 22:10:47 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 22:10:47 2015 +0200 dmeventd: fix missing '!' During recent code changes '!' was badly converted. --- .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index aebe5f9..d5575f3 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -201,7 +201,7 @@ int register_device(const char *device, { struct dso_state *state; - if (dmeventd_lvm2_init_with_pool("snapshot_state", state)) + if (!dmeventd_lvm2_init_with_pool("snapshot_state", state)) goto_bad; if (!dmeventd_lvm2_command(state->mem, state->cmd_lvextend,

1 0

master - makefiles: older gcc needs hint with rpath
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=a91fbe9d27a79b4b… Commit: a91fbe9d27a79b4be7fad72fc7a1ba2a976ecd41 Parent: ccc39be053b9a186abc6a020f516ff5f425981d6 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 22:02:17 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 22:02:17 2015 +0200 makefiles: older gcc needs hint with rpath gcc 4.3 seems not to be able to find linked library without specifying -rpath to linker (plain -L) is not enough. --- scripts/Makefile.in | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/scripts/Makefile.in b/scripts/Makefile.in index f7f672d..bfa0fad 100644 --- a/scripts/Makefile.in +++ b/scripts/Makefile.in @@ -25,6 +25,9 @@ include $(top_builddir)/make.tmpl ifeq ("@APPLIB@", "yes") DEPLIBS += $(top_builddir)/liblvm/liblvm2app.so $(top_builddir)/libdm/libdevmapper.so LDFLAGS += -L$(top_builddir)/liblvm +ifeq ("@DMEVENTD@", "yes") + LDFLAGS += -Wl,-rpath,$(top_builddir)/daemons/dmeventd +endif LVMLIBS = @LVM2APP_LIB@ -ldevmapper endif

1 0

master - dmeventd: compilable without DEBUG CFLAG
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=ccc39be053b9a186… Commit: ccc39be053b9a186abc6a020f516ff5f425981d6 Parent: 0cf787a377548394f4588b671c2890ee1a45098b Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 20:57:59 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 20:59:35 2015 +0200 dmeventd: compilable without DEBUG CFLAG Missed compilability without DEBUG. --- daemons/dmeventd/dmeventd.c | 9 ++++++--- 1 files changed, 6 insertions(+), 3 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index ba7d01c..56c504a 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -112,8 +112,6 @@ static int _restart = 0; static char **_initial_registrations = 0; /* FIXME Make configurable at runtime */ -#ifdef DEBUG -# define DEBUGLOG log_debug __attribute__((format(printf, 4, 5))) static void _dmeventd_log(int level, const char *file, int line, const char *format, ...) @@ -124,6 +122,8 @@ static void _dmeventd_log(int level, const char *file, int line, va_end(ap); } +#ifdef DEBUG +# define DEBUGLOG log_debug static const char *decode_cmd(uint32_t cmd) { switch (cmd) { @@ -1535,8 +1535,9 @@ static void _process_request(struct dm_event_fifos *fifos) { int die; struct dm_event_daemon_message msg = { 0 }; +#ifdef DEBUG int cmd; - +#endif /* * Read the request from the client (client_read, client_write * give true on success and false on failure). @@ -1544,7 +1545,9 @@ static void _process_request(struct dm_event_fifos *fifos) if (!_client_read(fifos, &msg)) return; +#ifdef DEBUG cmd = msg.cmd; +#endif DEBUGLOG(">>> CMD:%s (0x%x) processing...", decode_cmd(cmd), cmd); die = (msg.cmd == DM_EVENT_CMD_DIE) ? 1 : 0;

1 0

master - Revert "log: no file for external logging"
by Alasdair Kergon 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=0cf787a377548394… Commit: 0cf787a377548394f4588b671c2890ee1a45098b Parent: acc70de43968961e9a71612cda04642e4fbfd71f Author: Alasdair G Kergon <agk(a)redhat.com> AuthorDate: Tue Oct 13 15:31:57 2015 +0100 Committer: Alasdair G Kergon <agk(a)redhat.com> CommitterDate: Tue Oct 13 15:31:57 2015 +0100 Revert "log: no file for external logging" This reverts commit 1b1c01a27b359f8e91c3c9e08684c435d8fd51c2. This caused messages to get dropped instead of logged into the log file. (The log file and log function are independent at the moment.) --- WHATS_NEW | 1 - lib/log/log.c | 4 ---- 2 files changed, 0 insertions(+), 5 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 7507b9c..c089228 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,6 +1,5 @@ Version 2.02.133 - ====================================== - Avoid creation of log file when logging function is set. Do not change logging in lvm2 library when it's already set. Check for enough space in thin-pool in command before creating new thin. Make libblkid detect all copies of the same signature if use_blkid_wiping=1. diff --git a/lib/log/log.c b/lib/log/log.c index 2f1be72..c44c3d4 100644 --- a/lib/log/log.c +++ b/lib/log/log.c @@ -71,10 +71,6 @@ void init_log_file(const char *log_file, int append) int i = 0; _log_file_path[0] = '\0'; - - if (_lvm2_log_fn) - return; /* No log file, when log function is set */ - if ((env = getenv("LVM_LOG_FILE_EPOCH"))) { while (isalpha(env[i]) && i < 32) /* Up to 32 alphas */ i++;

1 0

master - tests: more extend testing
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=acc70de43968961e… Commit: acc70de43968961e9a71612cda04642e4fbfd71f Parent: cf1c2da836223de64daacd94f0fd40a96f6cd46f Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 14:33:24 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:02:21 2015 +0200 tests: more extend testing --- test/shell/lvextend-thin-metadata-dmeventd.sh | 51 +++++++++++++++++++++++-- 1 files changed, 47 insertions(+), 4 deletions(-) diff --git a/test/shell/lvextend-thin-metadata-dmeventd.sh b/test/shell/lvextend-thin-metadata-dmeventd.sh index ea081a2..9ecd5ae 100644 --- a/test/shell/lvextend-thin-metadata-dmeventd.sh +++ b/test/shell/lvextend-thin-metadata-dmeventd.sh @@ -37,14 +37,20 @@ wait_for_change_() { # Currently it expects 2MB thin metadata and 200MB data volume size # Argument specifies how many devices should be created. fake_metadata_() { - echo '<superblock uuid="" time="1" transaction="'$2'" data_block_size="128" nr_data_blocks="3200">' - for i in $(seq 1 $1) + echo '<superblock uuid="" time="0" transaction="'$2'" data_block_size="128" nr_data_blocks="3200">' + echo ' <device dev_id="1" mapped_blocks="0" transaction="0" creation_time="0" snap_time="0">' + echo ' </device>' + echo ' <device dev_id="2" mapped_blocks="0" transaction="0" creation_time="0" snap_time="0">' + echo ' </device>' + for i in $(seq 10 $1) do - echo ' <device dev_id="'$i'" mapped_blocks="37" transaction="0" creation_time="0" snap_time="1">' - echo ' <range_mapping origin_begin="0" data_begin="0" length="37" time="0"/>' + echo ' <device dev_id="'$i'" mapped_blocks="30" transaction="0" creation_time="0" snap_time="0">' + echo ' <range_mapping origin_begin="0" data_begin="0" length="29" time="0"/>' echo ' </device>' + set +x done echo "</superblock>" + set -x } test -n "$LVM_TEST_THIN_RESTORE_CMD" || LVM_TEST_THIN_RESTORE_CMD=$(which thin_restore) || skip @@ -99,4 +105,41 @@ vgchange -ay $vg pre=$(meta_percent_) wait_for_change_ $pre +lvchange -an $vg + +lvs -a $vg +# +fake_metadata_ 350 2 >data +lvchange -ay $vg/$lv1 +"$LVM_TEST_THIN_RESTORE_CMD" -i data -o "$DM_DEV_DIR/mapper/$vg-$lv1" + +lvs -a $vg +dmsetup table +lvconvert -y --chunksize 64k --thinpool $vg/pool --poolmetadata $vg/$lv1 +lvchange -ay $vg/pool $vg/$lv1 +lvs -a $vg + +lvcreate -s -Ky -n $lv2 $vg/thin +echo 2 >"$DM_DEV_DIR/mapper/$vg-$lv2" + +#lvchange -an $vg +#lvconvert -y --chunksize 64k --thinpool $vg/pool --poolmetadata $vg/$lv1 +#lvchange -ay $vg/$lv1 +#thin_dump "$DM_DEV_DIR/mapper/$vg-$lv1" +#exit + +# no more space for new thin LV +# +# TODO: +# though maybe 'lvcreate' itself should initiate resize - if dmeventd is not 'fast' enough +# deploying usage of threshold kernel module paramater would likely help as well +# for now it stops here: +not lvcreate -s -Ky -n $lv3 $vg/thin +lvs -a $vg + +pre=$(meta_percent_) +wait_for_change_ $pre + +lvs -a $vg + vgremove -f $vg

1 0

master - tests: wait for initial sync
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=cf1c2da836223de6… Commit: cf1c2da836223de64daacd94f0fd40a96f6cd46f Parent: c4cc5eabfe722daf88d19b60d58c616c31e57edd Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 11:21:40 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:02:21 2015 +0200 tests: wait for initial sync Raid should be in-sync ATM for any gaming. --- test/shell/lvconvert-repair-raid-dmeventd.sh | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/test/shell/lvconvert-repair-raid-dmeventd.sh b/test/shell/lvconvert-repair-raid-dmeventd.sh index df55449..a5a7a53 100644 --- a/test/shell/lvconvert-repair-raid-dmeventd.sh +++ b/test/shell/lvconvert-repair-raid-dmeventd.sh @@ -23,6 +23,7 @@ lvcreate -aey --type raid1 -m 3 --ignoremonitoring -L 1 -n 4way $vg lvchange --monitor y $vg/4way lvs -a -o all,lv_modules $vg lvdisplay --maps $vg +aux wait_for_sync $vg 4way aux disable_dev "$dev2" "$dev4" mkfs.ext3 "$DM_DEV_DIR/$vg/4way" sleep 10 # FIXME: need a "poll" utility, akin to "check"

1 0

master - tests: start logging dmeventd output
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=c4cc5eabfe722daf… Commit: c4cc5eabfe722daf88d19b60d58c616c31e57edd Parent: 8cc21354c2b4cb58223950b7e089eb28e602f0e8 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 09:44:08 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:02:21 2015 +0200 tests: start logging dmeventd output Log dmeventd output from test - keep output logging in test and avoid sending messages to syslog. --- test/lib/aux.sh | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/test/lib/aux.sh b/test/lib/aux.sh index 56316f7..8a25d87 100644 --- a/test/lib/aux.sh +++ b/test/lib/aux.sh @@ -70,7 +70,7 @@ prepare_dmeventd() { local run_valgrind= test "${LVM_VALGRIND_DMEVENTD:-0}" -eq 0 || run_valgrind="run_valgrind" - LVM_LOG_FILE_EPOCH=DMEVENTD $run_valgrind dmeventd -f "$@" & + LVM_LOG_FILE_EPOCH=DMEVENTD $run_valgrind dmeventd -fddddl "$@" >debug.log_DMEVENTD_out 2>&1 & echo $! > LOCAL_DMEVENTD # FIXME wait for pipe in /var/run instead

1 0

master - tests: better regex check
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=8cc21354c2b4cb58… Commit: 8cc21354c2b4cb58223950b7e089eb28e602f0e8 Parent: 2cb1f6eafeb72dcd49a7ed941d9c5400540284fa Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 9 21:30:37 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:02:20 2015 +0200 tests: better regex check Enhance check for matching prefix. --- test/lib/inittest.sh | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/test/lib/inittest.sh b/test/lib/inittest.sh index 4ed8d20..28bc4e8 100644 --- a/test/lib/inittest.sh +++ b/test/lib/inittest.sh @@ -30,7 +30,7 @@ COMMON_PREFIX="LVMTEST" PREFIX="${COMMON_PREFIX}$$" # Check we are not conflickting with some exiting setup -dmsetup table | not grep "$PREFIX" || die "DM table already has devices with prefix $PREFIX!" +dmsetup table | not grep "${PREFIX}[^0-9]" || die "DM table already has devices with prefix $PREFIX!" if test -z "$LVM_TEST_DIR"; then LVM_TEST_DIR=$TMPDIR; fi TESTDIR=$(mkdtemp "${LVM_TEST_DIR:-/tmp}" "$PREFIX.XXXXXXXXXX") || \

1 0

master - tests: show status in in_sync
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=2cb1f6eafeb72dcd… Commit: 2cb1f6eafeb72dcd49a7ed941d9c5400540284fa Parent: 93229184066a9b7726e3b572bca3f612ed2815fb Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Wed Oct 7 15:07:15 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:02:20 2015 +0200 tests: show status in in_sync Make visible actual status line when returing 'raid' is in sync. Useful to prove when kernel is wrong. --- test/lib/check.sh | 25 ++++++++++++++----------- 1 files changed, 14 insertions(+), 11 deletions(-) diff --git a/test/lib/check.sh b/test/lib/check.sh index b47bc8b..74a6eb6 100644 --- a/test/lib/check.sh +++ b/test/lib/check.sh @@ -189,16 +189,17 @@ in_sync() { local lvm_name="$1/$2" local dm_name=$(echo $lvm_name | sed s:-:--: | sed s:/:-:) - if ! a=(`dmsetup status $dm_name`); then + a=( $(dmsetup status $dm_name) ) || \ die "Unable to get sync status of $1" - elif [ ${a[2]} = "snapshot-origin" ]; then - if ! a=(`dmsetup status ${dm_name}-real`); then + + if [ ${a[2]} = "snapshot-origin" ]; then + a=( $(dmsetup status ${dm_name}-real) ) || \ die "Unable to get sync status of $1" - fi snap=": under snapshot" fi - if [ ${a[2]} = "raid" ]; then + case ${a[2]} in + "raid") # 6th argument is the sync ratio for RAID idx=6 type=${a[3]} @@ -206,13 +207,16 @@ in_sync() { echo "$lvm_name ($type$snap) is not in-sync" return 1 fi - elif [ ${a[2]} = "mirror" ]; then + ;; + "mirror") # 4th Arg tells us how far to the sync ratio idx=$((${a[3]} + 4)) type=${a[2]} - else + ;; + *) die "Unable to get sync ratio for target type '${a[2]}'" - fi + ;; + esac b=( $(echo ${a[$idx]} | sed s:/:' ':) ) @@ -221,11 +225,10 @@ in_sync() { return 1 fi - if [[ ${a[$(($idx - 1))]} =~ a ]]; then + [[ ${a[$(($idx - 1))]} =~ a ]] && \ die "$lvm_name ($type$snap) in-sync, but 'a' characters in health status" - fi - echo "$lvm_name ($type$snap) is in-sync" + echo "$lvm_name ($type$snap) is in-sync \"${a[@]}\"" } active() {

1 0

master - makefiles: src dirs have depth 5
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=93229184066a9b77… Commit: 93229184066a9b7726e3b572bca3f612ed2815fb Parent: 83f00e91567d387caf02da9cd3791c3fef85c80d Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 12 00:10:37 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:02:20 2015 +0200 makefiles: src dirs have depth 5 For tags generation scan depth 5. --- Makefile.in | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/Makefile.in b/Makefile.in index 68fbc79..e4d9937 100644 --- a/Makefile.in +++ b/Makefile.in @@ -234,7 +234,7 @@ ifneq ($(shell which ctags),) .PHONY: tags tags: test -z "$(shell find $(top_srcdir) -type f -name '*.[ch]' -newer tags 2>/dev/null | head -1)" || $(RM) tags - test -f tags || find $(top_srcdir) -maxdepth 4 -type f -name '*.[ch]' -exec ctags -a '{}' + + test -f tags || find $(top_srcdir) -maxdepth 5 -type f -name '*.[ch]' -exec ctags -a '{}' + CLEAN_TARGETS += tags endif

1 0

master - makefiles: drop explicit linking
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=83f00e91567d387c… Commit: 83f00e91567d387caf02da9cd3791c3fef85c80d Parent: 4f9e7f692ea1dc6370ee9978b34c4255d28b3c3c Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 9 21:33:59 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:02:20 2015 +0200 makefiles: drop explicit linking Nothing in /sscripts needs devmapper-event library so drop its linking. --- scripts/Makefile.in | 5 ----- 1 files changed, 0 insertions(+), 5 deletions(-) diff --git a/scripts/Makefile.in b/scripts/Makefile.in index dabe2ee..f7f672d 100644 --- a/scripts/Makefile.in +++ b/scripts/Makefile.in @@ -28,11 +28,6 @@ ifeq ("@APPLIB@", "yes") LVMLIBS = @LVM2APP_LIB@ -ldevmapper endif -ifeq ("@DMEVENTD@", "yes") - DEPLIBS += $(top_builddir)/daemons/dmeventd/libdevmapper-event.so - LVMLIBS += -ldevmapper-event -endif - LVM_SCRIPTS = lvmdump.sh lvmconf.sh vgimportclone.sh DM_SCRIPTS =

1 0

master - makefiles: generate Local
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=4f9e7f692ea1dc63… Commit: 4f9e7f692ea1dc6370ee9978b34c4255d28b3c3c Parent: 4b586ad3c26be5eaaeff38d767d5f2421ee1151b Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 9 21:31:41 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:02:20 2015 +0200 makefiles: generate Local Use same exported symbols list for other generated symfile as it appeared to be the right way for libdm. Local { local: *; }; --- make.tmpl.in | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/make.tmpl.in b/make.tmpl.in index 1217278..f1d4fad 100644 --- a/make.tmpl.in +++ b/make.tmpl.in @@ -497,7 +497,9 @@ EXPORTED_SYMBOLS := $(wildcard $(srcdir)/.exported_symbols.Base $(srcdir)/.expor ifeq (,$(firstword $(EXPORTED_SYMBOLS))) set -e; (echo "Base {"; echo " global:";\ $(SED) "s/^/ /;s/$$/;/" $<;\ - echo " local:"; echo " *;"; echo "};") > $@ + echo "};";\ + echo "Local {"; echo " local:"; echo " *;"; echo "};";\ + ) > $@ else set -e;\ R=$$(sort $^ | uniq -u);\

1 0

master - dmeventd: rework raid plugin
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=4b586ad3c26be5ea… Commit: 4b586ad3c26be5eaaeff38d767d5f2421ee1151b Parent: 256e432e78b720b929a285f5809b1d875b01861a Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 11:37:14 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:02:19 2015 +0200 dmeventd: rework raid plugin Switch to per-thread mem-pool and lock only for lvm2 calls. Use libdm parser for raid status line. --- WHATS_NEW_DM | 1 + daemons/dmeventd/plugins/raid/dmeventd_raid.c | 116 ++++++------------------- 2 files changed, 29 insertions(+), 88 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 30ce2cd..dffcd34 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Use libdm status parsing and local mem raid dmeventd plugin. Use local mem pool and lock only lvm2 execution for mirror dmeventd plugin. Lock protect only lvm2 execution for snapshot and thin dmeventd plugin. Use local mempool for raid and mirror plugins. diff --git a/daemons/dmeventd/plugins/raid/dmeventd_raid.c b/daemons/dmeventd/plugins/raid/dmeventd_raid.c index 714832b..f00ffa4 100644 --- a/daemons/dmeventd/plugins/raid/dmeventd_raid.c +++ b/daemons/dmeventd/plugins/raid/dmeventd_raid.c @@ -27,112 +27,54 @@ DM_EVENT_LOG_FN("raid") /* FIXME Reformat to 80 char lines. */ -/* - * run_repair is a close copy to - * plugins/mirror/dmeventd_mirror.c:_remove_failed_devices() - */ -static int run_repair(const char *device) +static int _process_raid_event(struct dso_state *state, char *params, const char *device) { - int r; -#define CMD_SIZE 256 /* FIXME Use system restriction */ - char cmd_str[CMD_SIZE]; - - if (!dmeventd_lvm2_command(dmeventd_lvm2_pool(), cmd_str, sizeof(cmd_str), - "lvscan --cache", device)) - return -1; - - r = dmeventd_lvm2_run(cmd_str); - - if (!r) - log_info("Re-scan of RAID device %s failed.", device); - - if (!dmeventd_lvm2_command(dmeventd_lvm2_pool(), cmd_str, sizeof(cmd_str), - "lvconvert --config devices{ignore_suspended_devices=1} " - "--repair --use-policies", device)) - return -1; - - /* if repair goes OK, report success even if lvscan has failed */ - r = dmeventd_lvm2_run(cmd_str); - - if (!r) - log_info("Repair of RAID device %s failed.", device); - - return (r) ? 0 : -1; -} + struct dm_status_raid *status; + const char *d; -static int _process_raid_event(char *params, const char *device) -{ - int i, n, failure = 0; - char *p, *a[4]; - char *raid_type; - char *num_devices; - char *health_chars; - char *resync_ratio; - - /* - * RAID parms: <raid_type> <#raid_disks> \ - * <health chars> <resync ratio> - */ - if (!dm_split_words(params, 4, 0, a)) { + if (!dm_get_status_raid(state->mem, params, &status)) { log_error("Failed to process status line for %s.", device); - return -EINVAL; - } - raid_type = a[0]; - num_devices = a[1]; - health_chars = a[2]; - resync_ratio = a[3]; - - if (!(n = atoi(num_devices))) { - log_error("Failed to parse number of devices for %s: %s.", - device, num_devices); - return -EINVAL; + return 0; } - for (i = 0; i < n; i++) { - switch (health_chars[i]) { - case 'A': - /* Device is 'A'live and well */ - case 'a': - /* Device is 'a'live, but not yet in-sync */ - break; - case 'D': - log_error("Device #%d of %s array, %s, has failed.", - i, raid_type, device); - failure++; - break; - default: - /* Unhandled character returned from kernel */ - break; + if ((d = strchr(status->dev_health, 'D')) && !state->failed) { + log_error("Device #%d of %s array, %s, has failed.", + (int)(d - status->dev_health), + status->raid_type, device); + + state->failed = 1; + if (!dmeventd_lvm2_run_with_lock(state->cmd_lvscan)) + log_info("Re-scan of RAID device %s failed.", device); + + /* if repair goes OK, report success even if lvscan has failed */ + if (!dmeventd_lvm2_run_with_lock(state->cmd_lvconvert)) { + log_info("Repair of RAID device %s failed.", device); + dm_pool_free(state->mem, status); + return 0; } - if (failure) - return run_repair(device); + } else { + state->failed = 0; + log_info("%s array, %s, is %s in-sync.", + status->raid_type, device, + (status->insync_regions == status->total_regions) ? "now" : "not"); } - p = strstr(resync_ratio, "/"); - if (!p) { - log_error("Failed to parse resync_ratio for %s: %s.", - device, resync_ratio); - return -EINVAL; - } - p[0] = '\0'; - log_info("%s array, %s, is %s in-sync.", - raid_type, device, strcmp(resync_ratio, p+1) ? "not" : "now"); + dm_pool_free(state->mem, status); - return 0; + return 1; } void process_event(struct dm_task *dmt, enum dm_event_mask event __attribute__((unused)), void **user) { + struct dso_state *state = *user; void *next = NULL; uint64_t start, length; char *target_type = NULL; char *params; const char *device = dm_task_get_name(dmt); - dmeventd_lvm2_lock(); - do { next = dm_get_next_target(dmt, next, &start, &length, &target_type, &params); @@ -147,12 +89,10 @@ void process_event(struct dm_task *dmt, continue; } - if (_process_raid_event(params, device)) + if (!_process_raid_event(state, params, device)) log_error("Failed to process event for %s.", device); } while (next); - - dmeventd_lvm2_unlock(); } int register_device(const char *device,

1 0

master - dmeventd: less locking mirror
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=256e432e78b720b9… Commit: 256e432e78b720b929a285f5809b1d875b01861a Parent: 51ff7d5ed8a7afe4e986a08025da9ef294aea479 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 11:35:44 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:02:18 2015 +0200 dmeventd: less locking mirror Rework mirror plugin to use per-thread mem-pool. This allows to use locking only for lvm2 library calls. --- WHATS_NEW_DM | 1 + daemons/dmeventd/plugins/mirror/dmeventd_mirror.c | 33 ++++++--------------- 2 files changed, 10 insertions(+), 24 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index f838b98..30ce2cd 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Use local mem pool and lock only lvm2 execution for mirror dmeventd plugin. Lock protect only lvm2 execution for snapshot and thin dmeventd plugin. Use local mempool for raid and mirror plugins. Reworked thread initialization for dmeventd plugins. diff --git a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c index e8761d9..62ad331 100644 --- a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c +++ b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c @@ -131,47 +131,33 @@ out_parse: return ME_IGNORE; } -static int _remove_failed_devices(const char *device) +static int _remove_failed_devices(const char *cmd_lvscan, const char *cmd_lvconvert) { int r; -#define CMD_SIZE 256 /* FIXME Use system restriction */ - char cmd_str[CMD_SIZE]; - if (!dmeventd_lvm2_command(dmeventd_lvm2_pool(), cmd_str, sizeof(cmd_str), - "lvscan --cache", device)) - return -1; - - r = dmeventd_lvm2_run(cmd_str); - - if (!r) - log_info("Re-scan of mirror device %s failed.", device); - - if (!dmeventd_lvm2_command(dmeventd_lvm2_pool(), cmd_str, sizeof(cmd_str), - "lvconvert --config devices{ignore_suspended_devices=1} " - "--repair --use-policies", device)) - return -ENAMETOOLONG; /* FIXME Replace with generic error return - reason for failure has already got logged */ + if (!dmeventd_lvm2_run_with_lock(cmd_lvscan)) + log_info("Re-scan of mirrored device failed."); /* if repair goes OK, report success even if lvscan has failed */ - r = dmeventd_lvm2_run(cmd_str); + r = dmeventd_lvm2_run_with_lock(cmd_lvconvert); - log_info("Repair of mirrored device %s %s.", device, + log_info("Repair of mirrored device %s.", (r) ? "finished successfully" : "failed"); - return (r) ? 0 : -1; + return r; } void process_event(struct dm_task *dmt, enum dm_event_mask event __attribute__((unused)), void **user) { + struct dso_state *state = *user; void *next = NULL; uint64_t start, length; char *target_type = NULL; char *params; const char *device = dm_task_get_name(dmt); - dmeventd_lvm2_lock(); - do { next = dm_get_next_target(dmt, next, &start, &length, &target_type, &params); @@ -196,7 +182,8 @@ void process_event(struct dm_task *dmt, break; case ME_FAILURE: log_error("Device failure in %s.", device); - if (_remove_failed_devices(device)) + if (!_remove_failed_devices(state->cmd_lvscan, + state->cmd_lvconvert)) /* FIXME Why are all the error return codes unused? Get rid of them? */ log_error("Failed to remove faulty devices in %s.", device); @@ -213,8 +200,6 @@ void process_event(struct dm_task *dmt, log_info("Unknown event received."); } } while (next); - - dmeventd_lvm2_unlock(); } int register_device(const char *device,

1 0

master - dmeventd: less locking for snapshot and thin
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=51ff7d5ed8a7afe4… Commit: 51ff7d5ed8a7afe4e986a08025da9ef294aea479 Parent: 49e11102c78ddf74bead6020fc51cae0e78e2ad2 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 11:35:03 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 16:01:41 2015 +0200 dmeventd: less locking for snapshot and thin Use lvm lock for lvm library calls. dm functions in-use are 'thread-safe' since we use local per-thread mem pool. --- WHATS_NEW_DM | 1 + .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 20 +++++++++----------- daemons/dmeventd/plugins/thin/dmeventd_thin.c | 7 +------ 3 files changed, 11 insertions(+), 17 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index 902a485..f838b98 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Lock protect only lvm2 execution for snapshot and thin dmeventd plugin. Use local mempool for raid and mirror plugins. Reworked thread initialization for dmeventd plugins. Dmeventd handles snapshot overflow for now equally as invalid. diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index 56fb7c2..aebe5f9 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -80,7 +80,8 @@ static int _run(const char *cmd, ...) static int _extend(const char *cmd) { - return dmeventd_lvm2_run(cmd); + log_debug("Extending snapshot via %s.", cmd); + return dmeventd_lvm2_run_with_lock(cmd); } static void _umount(const char *device, int major, int minor) @@ -141,22 +142,21 @@ void process_event(struct dm_task *dmt, if (!state->percent_check) return; - dmeventd_lvm2_lock(); - dm_get_next_target(dmt, next, &start, &length, &target_type, &params); - if (!target_type) - goto out; + if (!target_type || strcmp(target_type, "snapshot")) { + log_error("Target %s is not snapshot.", target_type); + return; + } if (!dm_get_status_snapshot(state->mem, params, &status)) - goto out; + return; if (status->invalid || status->overflow) { struct dm_info info; log_error("Snapshot %s is lost.", device); if (dm_task_get_info(dmt, &info)) { - dmeventd_lvm2_unlock(); _umount(device, info.major, info.minor); - return; + goto_out; } /* else; too bad, but this is best-effort thing... */ } @@ -190,9 +190,7 @@ void process_event(struct dm_task *dmt, log_error("Failed to extend snapshot %s.", device); } out: - if (status) - dm_pool_free(state->mem, status); - dmeventd_lvm2_unlock(); + dm_pool_free(state->mem, status); } int register_device(const char *device, diff --git a/daemons/dmeventd/plugins/thin/dmeventd_thin.c b/daemons/dmeventd/plugins/thin/dmeventd_thin.c index 15af254..4f08d5b 100644 --- a/daemons/dmeventd/plugins/thin/dmeventd_thin.c +++ b/daemons/dmeventd/plugins/thin/dmeventd_thin.c @@ -143,7 +143,7 @@ static int _extend(struct dso_state *state) #if THIN_DEBUG log_info("dmeventd executes: %s.", state->cmd_str); #endif - return dmeventd_lvm2_run(state->cmd_str); + return dmeventd_lvm2_run_with_lock(state->cmd_str); } static int _run(const char *cmd, ...) @@ -224,8 +224,6 @@ static void _umount(struct dm_task *dmt, const char *device) if (!dm_task_get_info(dmt, &data.info)) return; - dmeventd_lvm2_unlock(); - if (!(data.minors = dm_bitset_create(NULL, MINORS))) { log_error("Failed to allocate bitset. Not unmounting %s.", device); goto out; @@ -244,7 +242,6 @@ static void _umount(struct dm_task *dmt, const char *device) out: if (data.minors) dm_bitset_destroy(data.minors); - dmeventd_lvm2_lock(); } void process_event(struct dm_task *dmt, @@ -265,8 +262,6 @@ void process_event(struct dm_task *dmt, if (!state->meta_percent_check && !state->data_percent_check) return; #endif - dmeventd_lvm2_lock(); - dm_get_next_target(dmt, next, &start, &length, &target_type, &params); if (!target_type || (strcmp(target_type, "thin-pool") != 0)) {

1 0

master - dmeventd: add local mempool for raid and mirror
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=49e11102c78ddf74… Commit: 49e11102c78ddf74bead6020fc51cae0e78e2ad2 Parent: a11cd2ca2d4edbf16c61f8066b1f745fa2fe7d4b Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 11:28:57 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:56:23 2015 +0200 dmeventd: add local mempool for raid and mirror Using local mempools allows to drop locks when such memory is needed. --- WHATS_NEW_DM | 1 + daemons/dmeventd/plugins/mirror/dmeventd_mirror.c | 35 +++++++++++++++++++-- daemons/dmeventd/plugins/raid/dmeventd_raid.c | 32 +++++++++++++++++-- 3 files changed, 62 insertions(+), 6 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index f041ed1..902a485 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Use local mempool for raid and mirror plugins. Reworked thread initialization for dmeventd plugins. Dmeventd handles snapshot overflow for now equally as invalid. Convert dmeventd to use common logging macro system from libdm. diff --git a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c index d418b08..e8761d9 100644 --- a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c +++ b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c @@ -23,6 +23,12 @@ #define ME_INSYNC 1 #define ME_FAILURE 2 +struct dso_state { + struct dm_pool *mem; + char cmd_lvscan[512]; + char cmd_lvconvert[512]; +}; + DM_EVENT_LOG_FN("mirr") static int _process_status_code(const char status_code, const char *dev_name, @@ -217,12 +223,33 @@ int register_device(const char *device, int minor __attribute__((unused)), void **user) { - if (!dmeventd_lvm2_init()) - return 0; + struct dso_state *state; + + if (!dmeventd_lvm2_init_with_pool("mirror_state", state)) + goto_bad; + + if (!dmeventd_lvm2_command(state->mem, state->cmd_lvscan, sizeof(state->cmd_lvscan), + "lvscan --cache", device)) { + dmeventd_lvm2_exit_with_pool(state); + goto_bad; + } + + if (!dmeventd_lvm2_command(state->mem, state->cmd_lvconvert, sizeof(state->cmd_lvconvert), + "lvconvert --config devices{ignore_suspended_devices=1} " + "--repair --use-policies", device)) { + dmeventd_lvm2_exit_with_pool(state); + goto_bad; + } + + *user = state; log_info("Monitoring mirror device %s for events.", device); return 1; +bad: + log_error("Failed to monitor mirror %s.", device); + + return 0; } int unregister_device(const char *device, @@ -231,9 +258,11 @@ int unregister_device(const char *device, int minor __attribute__((unused)), void **user) { + struct dso_state *state = *user; + + dmeventd_lvm2_exit_with_pool(state); log_info("No longer monitoring mirror device %s for events.", device); - dmeventd_lvm2_exit(); return 1; } diff --git a/daemons/dmeventd/plugins/raid/dmeventd_raid.c b/daemons/dmeventd/plugins/raid/dmeventd_raid.c index 6522556..714832b 100644 --- a/daemons/dmeventd/plugins/raid/dmeventd_raid.c +++ b/daemons/dmeventd/plugins/raid/dmeventd_raid.c @@ -16,6 +16,13 @@ #include "dmeventd_lvm.h" #include "libdevmapper-event.h" +struct dso_state { + struct dm_pool *mem; + char cmd_lvscan[512]; + char cmd_lvconvert[512]; + int failed; +}; + DM_EVENT_LOG_FN("raid") /* FIXME Reformat to 80 char lines. */ @@ -154,12 +161,29 @@ int register_device(const char *device, int minor __attribute__((unused)), void **user) { - if (!dmeventd_lvm2_init()) - return 0; + struct dso_state *state; + + if (!dmeventd_lvm2_init_with_pool("raid_state", state)) + goto_bad; + + if (!dmeventd_lvm2_command(state->mem, state->cmd_lvscan, sizeof(state->cmd_lvscan), + "lvscan --cache", device) || + !dmeventd_lvm2_command(state->mem, state->cmd_lvconvert, sizeof(state->cmd_lvconvert), + "lvconvert --config devices{ignore_suspended_devices=1} " + "--repair --use-policies", device)) { + dmeventd_lvm2_exit_with_pool(state); + goto_bad; + } + + *user = state; log_info("Monitoring RAID device %s for events.", device); return 1; +bad: + log_error("Failed to monitor RAID %s.", device); + + return 0; } int unregister_device(const char *device, @@ -168,9 +192,11 @@ int unregister_device(const char *device, int minor __attribute__((unused)), void **user) { + struct dso_state *state = *user; + + dmeventd_lvm2_exit_with_pool(state); log_info("No longer monitoring RAID device %s for events.", device); - dmeventd_lvm2_exit(); return 1; }

1 0

master - dmeventd: use new macros to init mempool
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=a11cd2ca2d4edbf1… Commit: a11cd2ca2d4edbf16c61f8066b1f745fa2fe7d4b Parent: f9926e7e6c102319575c186e8eef0012cb9c36cc Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 11:30:37 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:56:03 2015 +0200 dmeventd: use new macros to init mempool For thin and snapshot use new macros to simplify mempool init and destroy --- .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 33 +++++++------------ daemons/dmeventd/plugins/thin/dmeventd_thin.c | 19 +++-------- 2 files changed, 18 insertions(+), 34 deletions(-) diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index 8d5bce1..56fb7c2 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -32,7 +32,7 @@ struct dso_state { struct dm_pool *mem; int percent_check; uint64_t known_size; - char cmd_str[1024]; + char cmd_lvextend[512]; }; DM_EVENT_LOG_FN("snap") @@ -184,11 +184,11 @@ void process_event(struct dm_task *dmt, if (percent >= WARNING_THRESH) /* Print a warning to syslog. */ log_warn("WARNING: Snapshot %s is now %i%% full.", device, percent); + /* Try to extend the snapshot, in accord with user-set policies */ - if (!_extend(state->cmd_str)) + if (!_extend(state->cmd_lvextend)) log_error("Failed to extend snapshot %s.", device); } - out: if (status) dm_pool_free(state->mem, status); @@ -201,22 +201,18 @@ int register_device(const char *device, int minor __attribute__((unused)), void **user) { - struct dm_pool *statemem = NULL; struct dso_state *state; - if (!dmeventd_lvm2_init()) - goto out; - - if (!(statemem = dm_pool_create("snapshot_state", 512)) || - !(state = dm_pool_zalloc(statemem, sizeof(*state)))) - goto bad; + if (dmeventd_lvm2_init_with_pool("snapshot_state", state)) + goto_bad; - if (!dmeventd_lvm2_command(statemem, state->cmd_str, - sizeof(state->cmd_str), - "lvextend --use-policies", device)) - goto bad; + if (!dmeventd_lvm2_command(state->mem, state->cmd_lvextend, + sizeof(state->cmd_lvextend), + "lvextend --use-policies", device)) { + dmeventd_lvm2_exit_with_pool(state); + goto_bad; + } - state->mem = statemem; state->percent_check = CHECK_MINIMUM; *user = state; @@ -224,10 +220,6 @@ int register_device(const char *device, return 1; bad: - if (statemem) - dm_pool_destroy(statemem); - dmeventd_lvm2_exit(); -out: log_error("Failed to monitor snapshot %s.", device); return 0; @@ -241,9 +233,8 @@ int unregister_device(const char *device, { struct dso_state *state = *user; + dmeventd_lvm2_exit_with_pool(state); log_info("No longer monitoring snapshot %s.", device); - dm_pool_destroy(state->mem); - dmeventd_lvm2_exit(); return 1; } diff --git a/daemons/dmeventd/plugins/thin/dmeventd_thin.c b/daemons/dmeventd/plugins/thin/dmeventd_thin.c index d21e65b..15af254 100644 --- a/daemons/dmeventd/plugins/thin/dmeventd_thin.c +++ b/daemons/dmeventd/plugins/thin/dmeventd_thin.c @@ -351,25 +351,19 @@ int register_device(const char *device, int minor __attribute__((unused)), void **user) { - struct dm_pool *statemem = NULL; struct dso_state *state; - if (!dmeventd_lvm2_init()) - goto bad; + if (!dmeventd_lvm2_init_with_pool("thin_pool_state", state)) + goto_bad; - if (!(statemem = dm_pool_create("thin_pool_state", 2048)) || - !(state = dm_pool_zalloc(statemem, sizeof(*state))) || - !dmeventd_lvm2_command(statemem, state->cmd_str, + if (!dmeventd_lvm2_command(state->mem, state->cmd_str, sizeof(state->cmd_str), "lvextend --use-policies", device)) { - if (statemem) - dm_pool_destroy(statemem); - dmeventd_lvm2_exit(); - goto bad; + dmeventd_lvm2_exit_with_pool(state); + goto_bad; } - state->mem = statemem; state->metadata_percent_check = CHECK_MINIMUM; state->data_percent_check = CHECK_MINIMUM; *user = state; @@ -391,9 +385,8 @@ int unregister_device(const char *device, { struct dso_state *state = *user; + dmeventd_lvm2_exit_with_pool(state); log_info("No longer monitoring thin %s.", device); - dm_pool_destroy(state->mem); - dmeventd_lvm2_exit(); return 1; }

1 0

master - dmeventd: introduce macro for init
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=f9926e7e6c102319… Commit: f9926e7e6c102319575c186e8eef0012cb9c36cc Parent: 76ea01dd207ce185891418c3f8ffdcff14bf2672 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 12 11:37:36 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:56:03 2015 +0200 dmeventd: introduce macro for init Simplify commonly used pool creation. --- daemons/dmeventd/plugins/lvm2/dmeventd_lvm.h | 34 +++++++++++++++++++++++++- 1 files changed, 33 insertions(+), 1 deletions(-) diff --git a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.h b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.h index 5691c87..d193eda 100644 --- a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.h +++ b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.h @@ -1,5 +1,5 @@ /* - * Copyright (C) 2010 Red Hat, Inc. All rights reserved. + * Copyright (C) 2010-2015 Red Hat, Inc. All rights reserved. * * This file is part of LVM2. * @@ -39,4 +39,36 @@ struct dm_pool *dmeventd_lvm2_pool(void); int dmeventd_lvm2_command(struct dm_pool *mem, char *buffer, size_t size, const char *cmd, const char *device); +#define dmeventd_lvm2_run_with_lock(cmdline) \ + ({\ + int rc;\ + dmeventd_lvm2_lock();\ + rc = dmeventd_lvm2_run(cmdline);\ + dmeventd_lvm2_unlock();\ + rc;\ + }) + +#define dmeventd_lvm2_init_with_pool(name, st) \ + ({\ + struct dm_pool *mem;\ + st = NULL;\ + if (dmeventd_lvm2_init()) {\ + if ((mem = dm_pool_create(name, 2048)) &&\ + (st = dm_pool_zalloc(mem, sizeof(*st))))\ + st->mem = mem;\ + else {\ + if (mem)\ + dm_pool_destroy(mem);\ + dmeventd_lvm2_exit();\ + }\ + }\ + st;\ + }) + +#define dmeventd_lvm2_exit_with_pool(pool) \ + do {\ + dm_pool_destroy(pool->mem);\ + dmeventd_lvm2_exit();\ + } while(0) + #endif /* _DMEVENTD_LVMWRAP_H */

1 0

master - dmeventd: new initialization of plugin threads
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=76ea01dd207ce185… Commit: 76ea01dd207ce185891418c3f8ffdcff14bf2672 Parent: 362558cd66bc08f8a46ababef66f2df21e8bd6fc Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 09:43:09 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:55:05 2015 +0200 dmeventd: new initialization of plugin threads Rework thread creation code to better use resources. New code will not leak 'timeout' registered thread on error path. Also if the thread already exist - avoid creation of thread object and it's later destruction. If the race is noticed during adding new monitoring thread, such thread is put on cleanup list and -EEXIST is reported. --- WHATS_NEW_DM | 1 + daemons/dmeventd/dmeventd.c | 170 +++++++++++++++++++++--------------------- 2 files changed, 86 insertions(+), 85 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index efe8b37..f041ed1 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Reworked thread initialization for dmeventd plugins. Dmeventd handles snapshot overflow for now equally as invalid. Convert dmeventd to use common logging macro system from libdm. Return -ENOMEM when device registration fails instead of 0 (=success). diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 3e54f79..ba7d01c 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -990,91 +990,6 @@ static int _active(struct message_data *message_data) } /* - * Register for an event. - * - * Only one caller at a time here, because we use - * a FIFO and lock it against multiple accesses. - */ -static int _register_for_event(struct message_data *message_data) -{ - int ret = 0; - struct thread_status *thread, *thread_new = NULL; - struct dso_data *dso_data; - - if (!(dso_data = _lookup_dso(message_data)) && - !(dso_data = _load_dso(message_data))) { - stack; -#ifdef ELIBACC - ret = -ELIBACC; -#else - ret = -ENODEV; -#endif - goto out; - } - - /* Preallocate thread status struct to avoid deadlock. */ - if (!(thread_new = _alloc_thread_status(message_data, dso_data))) { - stack; - ret = -ENOMEM; - goto out; - } - - if (!_fill_device_data(thread_new)) { - stack; - ret = -ENODEV; - goto out; - } - - /* If creation of timeout thread fails (as it may), we fail - here completely. The client is responsible for either - retrying later or trying to register without timeout - events. However, if timeout thread cannot be started, it - usually means we are so starved on resources that we are - almost as good as dead already... */ - if ((thread_new->events & DM_EVENT_TIMEOUT) && - (ret = -_register_for_timeout(thread_new))) - goto out; - - _lock_mutex(); - if (!(thread = _lookup_thread_status(message_data))) { - _unlock_mutex(); - - if (!_do_register_device(thread_new)) { - ret = -ENOMEM; - goto_out; - } - - thread = thread_new; - thread_new = NULL; - - /* Try to create the monitoring thread for this device. */ - _lock_mutex(); - if ((ret = -_create_thread(thread))) { - _unlock_mutex(); - _do_unregister_device(thread); - _free_thread_status(thread); - goto out; - } - - LINK_THREAD(thread); - } - - /* Or event # into events bitfield. */ - thread->events |= message_data->events_field; - _unlock_mutex(); - - out: - /* - * Deallocate thread status after releasing - * the lock in case we haven't used it. - */ - if (thread_new) - _free_thread_status(thread_new); - - return ret; -} - -/* * Unregister for an event. * * Only one caller at a time here as with register_for_event(). @@ -1128,6 +1043,91 @@ static int _unregister_for_event(struct message_data *message_data) } /* + * Register for an event. + * + * Only one caller at a time here, because we use + * a FIFO and lock it against multiple accesses. + */ +static int _register_for_event(struct message_data *message_data) +{ + int ret = 0; + struct thread_status *thread; + struct dso_data *dso_data; + + if (!(dso_data = _lookup_dso(message_data)) && + !(dso_data = _load_dso(message_data))) { + stack; +#ifdef ELIBACC + ret = -ELIBACC; +#else + ret = -ENODEV; +#endif + return ret; + } + + _lock_mutex(); + + if ((thread = _lookup_thread_status(message_data))) + /* Or event # into events bitfield. */ + thread->events |= message_data->events_field; + + _unlock_mutex(); + + if (!thread) { + if (!(thread = _alloc_thread_status(message_data, dso_data))) { + stack; + return -ENOMEM; + } + + if (!_fill_device_data(thread)) { + ret = -ENODEV; + goto_out; + } + + if (!_do_register_device(thread)) { + ret = -ENOMEM; + goto_out; + } + + if ((ret = -_create_thread(thread))) { + _do_unregister_device(thread); + goto_out; + } + + _lock_mutex(); + if (_lookup_thread_status(message_data)) { + DEBUGLOG("Race, uuid already registered, marking Thr %x unused.", + (int)thread->thread); + thread->status = DM_THREAD_SHUTDOWN; + thread->events = 0; + LINK(thread, &_thread_registry_unused); + _unlock_mutex(); + ret = -EEXIST; /* race ? */ + goto_out; + } + + LINK_THREAD(thread); + _unlock_mutex(); + } + + /* If creation of timeout thread fails (as it may), we fail + here completely. The client is responsible for either + retrying later or trying to register without timeout + events. However, if timeout thread cannot be started, it + usually means we are so starved on resources that we are + almost as good as dead already... */ + if ((thread->events & DM_EVENT_TIMEOUT) && + (ret = -_register_for_timeout(thread))) + _unregister_for_event(message_data); + + return ret; +out: + _free_thread_status(thread); + + return ret; +} + +/* * Get registered device. * * Only one caller at a time here as with register_for_event().

1 0

master - cleanup: typo in comment
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=362558cd66bc08f8… Commit: 362558cd66bc08f8a46ababef66f2df21e8bd6fc Parent: 09a8479cb70ea2761ebd6988926017aa13fc5f5d Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 12 11:37:04 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:22:58 2015 +0200 cleanup: typo in comment --- libdm/libdm-targets.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/libdm/libdm-targets.c b/libdm/libdm-targets.c index f22ad9c..166396c 100644 --- a/libdm/libdm-targets.c +++ b/libdm/libdm-targets.c @@ -82,7 +82,7 @@ int dm_get_status_raid(struct dm_pool *mem, const char *params, return_0; if (!(s->raid_type = dm_pool_zalloc(mem, p - params))) - goto_bad; /* memory is freed went pool is destroyed */ + goto_bad; /* memory is freed when pool is destroyed */ if (!(s->dev_health = dm_pool_zalloc(mem, i + 1))) goto_bad;

1 0

master - cleanup: move system defines before structs
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=09a8479cb70ea276… Commit: 09a8479cb70ea2761ebd6988926017aa13fc5f5d Parent: 0a633750f1e43aa655e5e6e05269aeab8650f55d Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 12 11:35:33 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:22:58 2015 +0200 cleanup: move system defines before structs --- daemons/dmeventd/plugins/thin/dmeventd_thin.c | 22 ++++++++++------------ 1 files changed, 10 insertions(+), 12 deletions(-) diff --git a/daemons/dmeventd/plugins/thin/dmeventd_thin.c b/daemons/dmeventd/plugins/thin/dmeventd_thin.c index 0801adf..d21e65b 100644 --- a/daemons/dmeventd/plugins/thin/dmeventd_thin.c +++ b/daemons/dmeventd/plugins/thin/dmeventd_thin.c @@ -19,11 +19,19 @@ #include <sys/wait.h> #include <stdarg.h> -/* First warning when thin is 80% full. */ +/* TODO - move this mountinfo code into library to be reusable */ +#ifdef __linux__ +# include "kdev_t.h" +#else +# define MAJOR(x) major((x)) +# define MINOR(x) minor((x)) +#endif + +/* First warning when thin data or metadata is 80% full. */ #define WARNING_THRESH 80 /* Run a check every 5%. */ #define CHECK_STEP 5 -/* Do not bother checking thins less than 50% full. */ +/* Do not bother checking thin data or metadata is less than 50% full. */ #define CHECK_MINIMUM 50 #define UMOUNT_COMMAND "/bin/umount" @@ -39,16 +47,6 @@ struct dso_state { char cmd_str[1024]; }; - -/* TODO - move this mountinfo code into library to be reusable */ -#ifdef __linux__ -# include "kdev_t.h" -#else -# define MAJOR(x) major((x)) -# define MINOR(x) minor((x)) -# define MKDEV(x,y) makedev((x),(y)) -#endif - DM_EVENT_LOG_FN("thin") /* Get dependencies for device, and try to find matching device */

1 0

master - cleanup: avoid using private
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=0a633750f1e43aa6… Commit: 0a633750f1e43aa655e5e6e05269aeab8650f55d Parent: 0e2261dbd1d489dff158fce88a705a680aaaa3da Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 12 11:40:51 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:22:58 2015 +0200 cleanup: avoid using private Switch private to user. --- daemons/dmeventd/plugins/mirror/dmeventd_mirror.c | 6 +++--- daemons/dmeventd/plugins/raid/dmeventd_raid.c | 6 +++--- .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 12 ++++++------ daemons/dmeventd/plugins/thin/dmeventd_thin.c | 12 ++++++------ 4 files changed, 18 insertions(+), 18 deletions(-) diff --git a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c index c010794..d418b08 100644 --- a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c +++ b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c @@ -156,7 +156,7 @@ static int _remove_failed_devices(const char *device) void process_event(struct dm_task *dmt, enum dm_event_mask event __attribute__((unused)), - void **unused __attribute__((unused))) + void **user) { void *next = NULL; uint64_t start, length; @@ -215,7 +215,7 @@ int register_device(const char *device, const char *uuid __attribute__((unused)), int major __attribute__((unused)), int minor __attribute__((unused)), - void **unused __attribute__((unused))) + void **user) { if (!dmeventd_lvm2_init()) return 0; @@ -229,7 +229,7 @@ int unregister_device(const char *device, const char *uuid __attribute__((unused)), int major __attribute__((unused)), int minor __attribute__((unused)), - void **unused __attribute__((unused))) + void **user) { log_info("No longer monitoring mirror device %s for events.", device); diff --git a/daemons/dmeventd/plugins/raid/dmeventd_raid.c b/daemons/dmeventd/plugins/raid/dmeventd_raid.c index 822d7c4..6522556 100644 --- a/daemons/dmeventd/plugins/raid/dmeventd_raid.c +++ b/daemons/dmeventd/plugins/raid/dmeventd_raid.c @@ -116,7 +116,7 @@ static int _process_raid_event(char *params, const char *device) void process_event(struct dm_task *dmt, enum dm_event_mask event __attribute__((unused)), - void **unused __attribute__((unused))) + void **user) { void *next = NULL; uint64_t start, length; @@ -152,7 +152,7 @@ int register_device(const char *device, const char *uuid __attribute__((unused)), int major __attribute__((unused)), int minor __attribute__((unused)), - void **unused __attribute__((unused))) + void **user) { if (!dmeventd_lvm2_init()) return 0; @@ -166,7 +166,7 @@ int unregister_device(const char *device, const char *uuid __attribute__((unused)), int major __attribute__((unused)), int minor __attribute__((unused)), - void **unused __attribute__((unused))) + void **user) { log_info("No longer monitoring RAID device %s for events.", device); diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index ae8e308..8d5bce1 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -126,8 +126,9 @@ static void _umount(const char *device, int major, int minor) void process_event(struct dm_task *dmt, enum dm_event_mask event __attribute__((unused)), - void **private) + void **user) { + struct dso_state *state = *user; void *next = NULL; uint64_t start, length; char *target_type = NULL; @@ -135,7 +136,6 @@ void process_event(struct dm_task *dmt, struct dm_status_snapshot *status = NULL; const char *device = dm_task_get_name(dmt); int percent; - struct dso_state *state = *private; /* No longer monitoring, waiting for remove */ if (!state->percent_check) @@ -199,7 +199,7 @@ int register_device(const char *device, const char *uuid __attribute__((unused)), int major __attribute__((unused)), int minor __attribute__((unused)), - void **private) + void **user) { struct dm_pool *statemem = NULL; struct dso_state *state; @@ -218,7 +218,7 @@ int register_device(const char *device, state->mem = statemem; state->percent_check = CHECK_MINIMUM; - *private = state; + *user = state; log_info("Monitoring snapshot %s.", device); @@ -237,9 +237,9 @@ int unregister_device(const char *device, const char *uuid __attribute__((unused)), int major __attribute__((unused)), int minor __attribute__((unused)), - void **private) + void **user) { - struct dso_state *state = *private; + struct dso_state *state = *user; log_info("No longer monitoring snapshot %s.", device); dm_pool_destroy(state->mem); diff --git a/daemons/dmeventd/plugins/thin/dmeventd_thin.c b/daemons/dmeventd/plugins/thin/dmeventd_thin.c index f57eb9b..0801adf 100644 --- a/daemons/dmeventd/plugins/thin/dmeventd_thin.c +++ b/daemons/dmeventd/plugins/thin/dmeventd_thin.c @@ -251,11 +251,11 @@ out: void process_event(struct dm_task *dmt, enum dm_event_mask event __attribute__((unused)), - void **private) + void **user) { const char *device = dm_task_get_name(dmt); int percent; - struct dso_state *state = *private; + struct dso_state *state = *user; struct dm_status_thin_pool *tps = NULL; void *next = NULL; uint64_t start, length; @@ -351,7 +351,7 @@ int register_device(const char *device, const char *uuid __attribute__((unused)), int major __attribute__((unused)), int minor __attribute__((unused)), - void **private) + void **user) { struct dm_pool *statemem = NULL; struct dso_state *state; @@ -374,7 +374,7 @@ int register_device(const char *device, state->mem = statemem; state->metadata_percent_check = CHECK_MINIMUM; state->data_percent_check = CHECK_MINIMUM; - *private = state; + *user = state; log_info("Monitoring thin %s.", device); @@ -389,9 +389,9 @@ int unregister_device(const char *device, const char *uuid __attribute__((unused)), int major __attribute__((unused)), int minor __attribute__((unused)), - void **private) + void **user) { - struct dso_state *state = *private; + struct dso_state *state = *user; log_info("No longer monitoring thin %s.", device); dm_pool_destroy(state->mem);

1 0

master - cleanup: remove multilog
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=0e2261dbd1d489df… Commit: 0e2261dbd1d489dff158fce88a705a680aaaa3da Parent: 842a7a17e3c0ffc0467ef23a3b59e4e8af2c8d74 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 09:40:39 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:22:57 2015 +0200 cleanup: remove multilog --- daemons/dmeventd/dmeventd.c | 6 ------ 1 files changed, 0 insertions(+), 6 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 7ad30b3..3e54f79 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -2152,11 +2152,6 @@ int main(int argc, char *argv[]) _init_thread_signals(); - //multilog_clear_logging(); - //multilog_add_type(std_syslog, &logdata); - //multilog_init_verbose(std_syslog, _LOG_DEBUG); - //multilog_async(1); - pthread_mutex_init(&_global_mutex, NULL); if (!_systemd_activation && !_open_fifos(&fifos)) @@ -2191,7 +2186,6 @@ int main(int argc, char *argv[]) _cleanup_unused_threads(); } - pthread_mutex_destroy(&_global_mutex); log_notice("dmeventd shutting down.");

1 0

master - cleanup: always set nsec
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=842a7a17e3c0ffc0… Commit: 842a7a17e3c0ffc0467ef23a3b59e4e8af2c8d74 Parent: f4fb97c85082b964397a4e7930d7aa809ffd58f9 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 09:39:18 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:22:57 2015 +0200 cleanup: always set nsec --- daemons/dmeventd/dmeventd.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index e2daef4..7ad30b3 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -568,12 +568,12 @@ static void *_timeout_thread(void *unused __attribute__((unused))) time_t curr_time; DEBUGLOG("Timeout thread starting."); - timeout.tv_nsec = 0; pthread_cleanup_push(_exit_timeout, NULL); pthread_mutex_lock(&_timeout_mutex); while (!dm_list_empty(&_timeout_registry)) { timeout.tv_sec = 0; + timeout.tv_nsec = 0; curr_time = time(NULL); dm_list_iterate_items_gen(thread, &_timeout_registry, timeout_list) {

1 0

master - cleanup: more readable code
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=f4fb97c85082b964… Commit: f4fb97c85082b964397a4e7930d7aa809ffd58f9 Parent: 8b9533f38f32143e73979542161e62801438922a Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 9 21:37:12 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:22:57 2015 +0200 cleanup: more readable code --- lib/log/log.c | 5 +---- lib/log/lvm-logging.h | 4 ++-- 2 files changed, 3 insertions(+), 6 deletions(-) diff --git a/lib/log/log.c b/lib/log/log.c index 7f31361..2f1be72 100644 --- a/lib/log/log.c +++ b/lib/log/log.c @@ -49,10 +49,7 @@ static size_t _lvm_errmsg_len = 0; void init_log_fn(lvm2_log_fn_t log_fn) { - if (log_fn) - _lvm2_log_fn = log_fn; - else - _lvm2_log_fn = NULL; + _lvm2_log_fn = log_fn; } /* diff --git a/lib/log/lvm-logging.h b/lib/log/lvm-logging.h index 983da5b..84bf3f4 100644 --- a/lib/log/lvm-logging.h +++ b/lib/log/lvm-logging.h @@ -16,9 +16,9 @@ #ifndef _LVM_LOGGING_H #define _LVM_LOGGING_H +__attribute__ ((format(printf, 5, 6))) void print_log(int level, const char *file, int line, int dm_errno_or_class, - const char *format, ...) - __attribute__ ((format(printf, 5, 6))); + const char *format, ...); #define LOG_LINE(l, x...) \ print_log(l, __FILE__, __LINE__ , 0, ## x)

1 0

master - dmeventd: support logging on stdout
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=8b9533f38f32143e… Commit: 8b9533f38f32143e73979542161e62801438922a Parent: 903e9af1b2c1972a897b7dd61e6e463a03cad648 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 15:14:11 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:22:57 2015 +0200 dmeventd: support logging on stdout Add new supported option '-l' log to stdout/stderr. It has to be paired with '-f' (foreground run). --- daemons/dmeventd/dmeventd.c | 33 ++++++++++++++++++++++++++------- man/dmeventd.8.in | 7 +++++++ 2 files changed, 33 insertions(+), 7 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index f775fb9..e2daef4 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -104,7 +104,8 @@ static pthread_mutex_t _global_mutex; #define THREAD_STACK_SIZE (300*1024) -int dmeventd_debug = 0; +static int _debug_level = 0; +static int _use_syslog = 1; static int _systemd_activation = 0; static int _foreground = 0; static int _restart = 0; @@ -2051,10 +2052,11 @@ bad: static void _usage(char *prog, FILE *file) { fprintf(file, "Usage:\n" - "%s [-d [-d [-d]]] [-f] [-h] [-R] [-V] [-?]\n\n" + "%s [-d [-d [-d]]] [-f] [-h] [-l] [-R] [-V] [-?]\n\n" " -d Log debug messages to syslog (-d, -dd, -ddd)\n" " -f Don't fork, run in the foreground\n" " -h Show this help information\n" + " -l Log to stdout,stderr instead of syslog\n" " -? Show this help information on stderr\n" " -R Restart dmeventd\n" " -V Show version of dmeventd\n\n", prog); @@ -2075,7 +2077,7 @@ int main(int argc, char *argv[]) opterr = 0; optind = 0; - while ((opt = getopt(argc, argv, "?fhVdR")) != EOF) { + while ((opt = getopt(argc, argv, "?fhVdlR")) != EOF) { switch (opt) { case 'h': _usage(argv[0], stdout); @@ -2090,7 +2092,10 @@ int main(int argc, char *argv[]) _foreground++; break; case 'd': - dmeventd_debug++; + _debug_level++; + break; + case 'l': + _use_syslog = 0; break; case 'V': printf("dmeventd version: %s\n", DM_LIB_VERSION); @@ -2098,6 +2103,10 @@ int main(int argc, char *argv[]) } } + if (!_foreground && !_use_syslog) { + printf("WARNING: Ignoring logging to stdout, needs options -f\n"); + _use_syslog = 1; + } /* * Switch to C locale to avoid reading large locale-archive file * used by some glibc (on some distributions it takes over 100MB). @@ -2116,8 +2125,10 @@ int main(int argc, char *argv[]) if (!_foreground) _daemonize(); - openlog("dmeventd", LOG_PID, LOG_DAEMON); + if (_use_syslog) + openlog("dmeventd", LOG_PID, LOG_DAEMON); + dm_event_log_set(_debug_level, _use_syslog); dm_log_init(_dmeventd_log); (void) dm_prepare_selinux_context(DMEVENTD_PIDFILE, S_IFREG); @@ -2180,12 +2191,20 @@ int main(int argc, char *argv[]) _cleanup_unused_threads(); } - _exit_dm_lib(); pthread_mutex_destroy(&_global_mutex); log_notice("dmeventd shutting down."); - closelog(); + + if (close(fifos.client)) + log_sys_error("client close", fifos.client_path); + if (close(fifos.server)) + log_sys_error("server close", fifos.server_path); + + if (_use_syslog) + closelog(); + + _exit_dm_lib(); exit(EXIT_SUCCESS); } diff --git a/man/dmeventd.8.in b/man/dmeventd.8.in index 15e003b..aa62451 100644 --- a/man/dmeventd.8.in +++ b/man/dmeventd.8.in @@ -12,6 +12,7 @@ dmeventd \(em Device-mapper event daemon .RB [ \-d ]]] .RB [ \-f ] .RB [ \-h ] +.RB [ \-l ] .RB [ \-R ] .RB [ \-V ] .RB [ \-? ] @@ -81,6 +82,12 @@ Don't fork, run in the foreground. Show help information. . .HP +.BR \-l +.br +Log through stdout and stderr instead of syslog. +This option works only with option \-f, otherwise it is ignored. +. +.HP .BR \-? .br Show help information on stderr.

1 0

master - cleanup: static with _
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=903e9af1b2c1972a… Commit: 903e9af1b2c1972a897b7dd61e6e463a03cad648 Parent: e261af52eb26c22aebd62cdbd8bb79c195274ee9 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 8 11:59:11 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:16:56 2015 +0200 cleanup: static with _ Document ? is printing to stderr. --- daemons/dmeventd/dmeventd.c | 17 +++++++++-------- 1 files changed, 9 insertions(+), 8 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index e34c4c1..f775fb9 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -931,7 +931,7 @@ static int _lookup_symbol(void *dl, void **symbol, const char *name) return 0; } -static int lookup_symbols(void *dl, struct dso_data *data) +static int _lookup_symbols(void *dl, struct dso_data *data) { return _lookup_symbol(dl, (void *) &data->process_event, "process_event") && @@ -962,7 +962,7 @@ static struct dso_data *_load_dso(struct message_data *data) return NULL; } - if (!(lookup_symbols(dl, ret))) { + if (!(_lookup_symbols(dl, ret))) { _free_dso_data(ret); dlclose(dl); return NULL; @@ -1931,7 +1931,7 @@ static int _reinstate_registrations(struct dm_event_fifos *fifos) return 1; } -static void restart(void) +static void _restart_dmeventd(void) { struct dm_event_fifos fifos = { .server = -1, @@ -2048,13 +2048,14 @@ bad: exit(EXIT_FAILURE); } -static void usage(char *prog, FILE *file) +static void _usage(char *prog, FILE *file) { fprintf(file, "Usage:\n" "%s [-d [-d [-d]]] [-f] [-h] [-R] [-V] [-?]\n\n" " -d Log debug messages to syslog (-d, -dd, -ddd)\n" " -f Don't fork, run in the foreground\n" - " -h -? Show this help information\n" + " -h Show this help information\n" + " -? Show this help information on stderr\n" " -R Restart dmeventd\n" " -V Show version of dmeventd\n\n", prog); } @@ -2077,10 +2078,10 @@ int main(int argc, char *argv[]) while ((opt = getopt(argc, argv, "?fhVdR")) != EOF) { switch (opt) { case 'h': - usage(argv[0], stdout); + _usage(argv[0], stdout); exit(EXIT_SUCCESS); case '?': - usage(argv[0], stderr); + _usage(argv[0], stderr); exit(EXIT_SUCCESS); case 'R': _restart++; @@ -2106,7 +2107,7 @@ int main(int argc, char *argv[]) perror("Cannot set LC_ALL to C"); if (_restart) - restart(); + _restart_dmeventd(); #ifdef __linux__ _systemd_activation = _systemd_handover(&fifos);

1 0

master - dmeventd: handle snapshot overflow
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=e261af52eb26c22a… Commit: e261af52eb26c22aebd62cdbd8bb79c195274ee9 Parent: 3f03d46fc13be2274b3c29d4fc97a3306bc33eda Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 9 21:41:50 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:15:44 2015 +0200 dmeventd: handle snapshot overflow When snapshot reports overflow, handle it in the same way as Invalid. Until better ideas are implemented. --- WHATS_NEW_DM | 1 + .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 4 ++-- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index ff4ac22..efe8b37 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Dmeventd handles snapshot overflow for now equally as invalid. Convert dmeventd to use common logging macro system from libdm. Return -ENOMEM when device registration fails instead of 0 (=success). Enforce writethrough mode for cleaner policy. diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index ffe6666..ae8e308 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -150,7 +150,7 @@ void process_event(struct dm_task *dmt, if (!dm_get_status_snapshot(state->mem, params, &status)) goto out; - if (status->invalid) { + if (status->invalid || status->overflow) { struct dm_info info; log_error("Snapshot %s is lost.", device); if (dm_task_get_info(dmt, &info)) { @@ -170,7 +170,7 @@ void process_event(struct dm_task *dmt, * If the snapshot has been invalidated or we failed to parse * the status string. Report the full status string to syslog. */ - if (status->invalid || !status->total_sectors) { + if (status->invalid || status->overflow || !status->total_sectors) { log_error("Snapshot %s changed state to: %s.", device, params); state->percent_check = 0; goto out;

1 0

master - man: dmeventd update
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=3f03d46fc13be227… Commit: 3f03d46fc13be2274b3c29d4fc97a3306bc33eda Parent: 0e2721030857088803bc71ccc27644b7b95241ca Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 12 11:35:53 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:15:43 2015 +0200 man: dmeventd update --- man/dmeventd.8.in | 92 +++++++++++++++++++++++++++++++++++++--------------- 1 files changed, 65 insertions(+), 27 deletions(-) diff --git a/man/dmeventd.8.in b/man/dmeventd.8.in index 77bbafb..15e003b 100644 --- a/man/dmeventd.8.in +++ b/man/dmeventd.8.in @@ -1,7 +1,11 @@ .TH DMEVENTD 8 "DM TOOLS #VERSION#" "Red Hat Inc" \" -*- nroff -*- +. .SH NAME +. dmeventd \(em Device-mapper event daemon +. .SH SYNOPSIS +. .B dmeventd .RB [ \-d .RB [ \-d @@ -11,33 +15,53 @@ dmeventd \(em Device-mapper event daemon .RB [ \-R ] .RB [ \-V ] .RB [ \-? ] +. .SH DESCRIPTION +. dmeventd is the event monitoring daemon for device-mapper devices. Library plugins can register and carry out actions triggered when particular events occur. -.SH -LVM PLUGINS -.TP -.I Mirror -Attempts to handle device failure automatically. See \fBlvm.conf\fP(5). -.TP -.I Raid -Attempts to handle device failure automatically. See \fBlvm.conf\fP(5). -.TP -.I Snapshot +. +.SH LVM PLUGINS +. +.HP +.IR Mirror +.br +Attempts to handle device failure automatically. See +.BR lvm.conf (5). +. +.HP +.IR Raid +.br +Attempts to handle device failure automatically. See +.BR lvm.conf (5). +. +.HP +.IR Snapshot +.br Monitors how full a snapshot is becoming and emits a warning to syslog when it exceeds 80% full. The warning is repeated when 85%, 90% and 95% of the snapshot is filled. -See \fBlvm.conf\fP(5). -.TP -.I Thin -Monitors how full a thin pool is becoming and emits a warning to -syslog when it exceeds 80% full. +See +.BR lvm.conf (5). +Snapshot which runs out of space gets invalid and when it is mounted, +it gets umounted if possible. +. +.HP +.IR Thin +.br +Monitors how full a thin pool data and metadata is becoming and emits +a warning to syslog when it exceeds 80% full. The warning is repeated when 85%, 90% and 95% of the thin pool is filled. -See \fBlvm.conf\fP(5). +See +.BR lvm.conf (5). +If the thin-pool runs out of space, thin volumes are umounted if possible. +. .SH OPTIONS -.TP -.B \-d +. +.HP +.BR \-d +.br Repeat from 1 to 3 times ( .BR \-d , .BR \-dd , @@ -45,21 +69,35 @@ Repeat from 1 to 3 times ( ) to increase the detail of debug messages sent to syslog. Each extra d adds more debugging information. -.TP -.B \-f +. +.HP +.BR \-f +.br Don't fork, run in the foreground. -.TP -.BR \-h ", " \-? +. +.HP +.BR \-h +.br Show help information. -.TP -.B \-R +. +.HP +.BR \-? +.br +Show help information on stderr. +. +.HP +.BR \-R +.br Replace a running dmeventd instance. The running dmeventd must be version 2.02.77 or newer. The new dmeventd instance will obtain a list of devices and events to monitor from the currently running daemon. -.TP -.B \-V +. +.HP +.BR \-V +.br Show version of dmeventd. - +. .SH SEE ALSO +. .BR lvm (8), .BR lvm.conf (5)

1 0

master - dmeventd: add more debug messages
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=0e27210308570888… Commit: 0e2721030857088803bc71ccc27644b7b95241ca Parent: 13086c2523a00bc8501113f56094312bda4881e1 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sat Oct 10 16:36:18 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:15:43 2015 +0200 dmeventd: add more debug messages --- daemons/dmeventd/dmeventd.c | 15 ++++++++++++--- daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c | 4 ++++ .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 1 + 3 files changed, 17 insertions(+), 3 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 0acb919..e34c4c1 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -874,7 +874,7 @@ static void *_monitor_thread(void *arg) } } - DEBUGLOG("Finished _monitor_thread"); + DEBUGLOG("Finished _monitor_thread."); pthread_cleanup_pop(1); return NULL; @@ -1120,6 +1120,8 @@ static int _unregister_for_event(struct message_data *message_data) } _unlock_mutex(); + DEBUGLOG("Unregistered uuid:%s.", thread->device.uuid); + out: return ret; } @@ -1147,6 +1149,7 @@ static int _registered_device(struct message_data *message_data, return -ENOMEM; msg->size = (uint32_t) r; + DEBUGLOG("Registered %s.", msg->data); return 0; } @@ -1154,6 +1157,7 @@ static int _registered_device(struct message_data *message_data, static int _want_registered_device(char *dso_name, char *device_uuid, struct thread_status *thread) { + DEBUGLOG("Looking for dso:%s uuid:%s.", dso_name, device_uuid); /* If DSO names and device paths are equal. */ if (dso_name && device_uuid) return !strcmp(dso_name, thread->dso_data->dso_name) && @@ -1175,6 +1179,9 @@ static int _get_registered_dev(struct message_data *message_data, int next) struct thread_status *thread, *hit = NULL; int ret = -ENOENT; + DEBUGLOG("Get%s dso:%s uuid:%s.", next ? "" : "Next", + message_data->dso_name, + message_data->device_uuid); _lock_mutex(); /* Iterate list of threads checking if we want a particular one. */ @@ -1205,8 +1212,10 @@ static int _get_registered_dev(struct message_data *message_data, int next) goto reg; } - if (!hit) + if (!hit) { + DEBUGLOG("Get%s not registered", next ? "" : "Next"); goto out; + } while (1) { if (dm_list_end(&_thread_registry, &thread->list)) @@ -1613,9 +1622,9 @@ static void _cleanup_unused_threads(void) } if (thread->status == DM_THREAD_DONE) { - DEBUGLOG("Destroying Thr %x.", (int)thread->thread); dm_list_del(l); _unlock_mutex(); + DEBUGLOG("Destroying Thr %x.", (int)thread->thread); join_ret = pthread_join(thread->thread, NULL); _free_thread_status(thread); _lock_mutex(); diff --git a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c index 0ba3694..e3d8321 100644 --- a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c +++ b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c @@ -74,6 +74,7 @@ int dmeventd_lvm2_init(void) lvm2_disable_dmeventd_monitoring(_lvm_handle); /* FIXME Temporary: move to dmeventd core */ lvm2_run(_lvm_handle, "_memlock_inc"); + log_debug("lvm plugin initilized."); } _register_count++; @@ -89,14 +90,17 @@ void dmeventd_lvm2_exit(void) pthread_mutex_lock(&_register_mutex); if (!--_register_count) { + log_debug("lvm plugin shuting down."); lvm2_run(_lvm_handle, "_memlock_dec"); dm_pool_destroy(_mem_pool); _mem_pool = NULL; + log_debug("lvm plugin exiting."); lvm2_exit(_lvm_handle); _lvm_handle = NULL; } pthread_mutex_unlock(&_register_mutex); + log_debug("lvm plugin exited."); } struct dm_pool *dmeventd_lvm2_pool(void) diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index f78b8f7..ffe6666 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -152,6 +152,7 @@ void process_event(struct dm_task *dmt, if (status->invalid) { struct dm_info info; + log_error("Snapshot %s is lost.", device); if (dm_task_get_info(dmt, &info)) { dmeventd_lvm2_unlock(); _umount(device, info.major, info.minor);

1 0

master - dmeventd: drop changing logging
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=13086c2523a00bc8… Commit: 13086c2523a00bc8501113f56094312bda4881e1 Parent: 42a9c8b4a659a44bef1b3150eb6359d562938ca3 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 9 22:08:01 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:15:43 2015 +0200 dmeventd: drop changing logging As we now use 'unified' logging macro system - we no longer need to protect from change of logging function pointer - it's set once at the start of dmeventd and not change anymore (as lvm2 library no longer interferers here). --- daemons/dmeventd/dmeventd.c | 10 ---------- 1 files changed, 0 insertions(+), 10 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 6f25379..0acb919 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -648,7 +648,6 @@ static sigset_t _unblock_sigalrm(void) /* Wait on a device until an event occurs. */ static int _event_wait(struct thread_status *thread, struct dm_task **task) { - static unsigned _in_event_counter = 0; sigset_t set; int ret = DM_WAIT_RETRY; struct dm_task *dmt; @@ -667,16 +666,11 @@ static int _event_wait(struct thread_status *thread, struct dm_task **task) !dm_task_set_event_nr(dmt, thread->event_nr)) goto out; - _lock_mutex(); /* * Check if there are already some waiting events, * in this case the logging is unmodified. * TODO: audit libdm thread usage */ - if (!_in_event_counter++) - dm_log_init(_dmeventd_log); - _unlock_mutex(); - DEBUGLOG("Starting waitevent task for %s", thread->device.uuid); /* * This is so that you can break out of waiting on an event, @@ -708,10 +702,6 @@ static int _event_wait(struct thread_status *thread, struct dm_task **task) DEBUGLOG("Completed waitevent task for %s", thread->device.uuid); pthread_sigmask(SIG_SETMASK, &set, NULL); - _lock_mutex(); - if (--_in_event_counter == 0) - dm_log_init(NULL); - _unlock_mutex(); out: if (ret == DM_WAIT_FATAL || ret == DM_WAIT_RETRY) {

1 0

master - dmeventd: enable new logging
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=42a9c8b4a659a44b… Commit: 42a9c8b4a659a44bef1b3150eb6359d562938ca3 Parent: e50583d7219e01228d697f91a16e91ed1f5efe86 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 15:13:28 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:13:28 2015 +0200 dmeventd: enable new logging --- daemons/dmeventd/dmeventd.c | 3 +-- 1 files changed, 1 insertions(+), 2 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 26609db..6f25379 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -674,7 +674,7 @@ static int _event_wait(struct thread_status *thread, struct dm_task **task) * TODO: audit libdm thread usage */ if (!_in_event_counter++) - dm_log_init(_no_intr_log); + dm_log_init(_dmeventd_log); _unlock_mutex(); DEBUGLOG("Starting waitevent task for %s", thread->device.uuid); @@ -2118,7 +2118,6 @@ int main(int argc, char *argv[]) openlog("dmeventd", LOG_PID, LOG_DAEMON); - dm_event_log_set(_debug_level, _use_syslog); dm_log_init(_dmeventd_log); (void) dm_prepare_selinux_context(DMEVENTD_PIDFILE, S_IFREG);

1 0

master - dmeventd: enable new logging
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=e50583d7219e0122… Commit: e50583d7219e01228d697f91a16e91ed1f5efe86 Parent: c90363b585a41de2a65872f9be359db22ee64325 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sat Oct 10 18:32:53 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:09:35 2015 +0200 dmeventd: enable new logging --- daemons/dmeventd/dmeventd.c | 40 +++++++++++----------------------------- 1 files changed, 11 insertions(+), 29 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 581ce18..26609db 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -16,14 +16,13 @@ * dmeventd - dm event daemon to monitor active mapped devices */ -#include "tool.h" - -//#include "libmultilog.h" #include "dm-logging.h" #include "libdevmapper-event.h" #include "dmeventd.h" +#include "tool.h" + #include <dlfcn.h> #include <pthread.h> #include <sys/file.h> @@ -61,7 +60,6 @@ #endif -/* FIXME We use syslog for now, because multilog is not yet implemented */ #include <syslog.h> static volatile sig_atomic_t _exit_now = 0; /* set to '1' when signal is given to exit */ @@ -114,15 +112,14 @@ static char **_initial_registrations = 0; /* FIXME Make configurable at runtime */ #ifdef DEBUG -# define DEBUGLOG(fmt, args...) debuglog("[Thr %x]: " fmt, (int)pthread_self(), ## args) -void debuglog(const char *fmt, ... ) __attribute__ ((format(printf, 1, 2))); - -void debuglog(const char *fmt, ...) +# define DEBUGLOG log_debug +__attribute__((format(printf, 4, 5))) +static void _dmeventd_log(int level, const char *file, int line, + const char *format, ...) { va_list ap; - - va_start(ap, fmt); - vsyslog(LOG_DEBUG, fmt, ap); + va_start(ap, format); + dm_event_log("dm", level, file, line, 0, format, ap); va_end(ap); } @@ -634,24 +631,6 @@ static void _unregister_for_timeout(struct thread_status *thread) pthread_mutex_unlock(&_timeout_mutex); } -__attribute__((format(printf, 4, 5))) -static void _no_intr_log(int level, const char *file, int line, - const char *f, ...) -{ - va_list ap; - - if (errno == EINTR) - return; - if (level > _LOG_WARN) - return; - - va_start(ap, f); - vfprintf((level < _LOG_WARN) ? stderr : stdout, f, ap); - va_end(ap); - - fputc('\n', (level < _LOG_WARN) ? stderr : stdout); -} - static sigset_t _unblock_sigalrm(void) { sigset_t set, old; @@ -2139,6 +2118,9 @@ int main(int argc, char *argv[]) openlog("dmeventd", LOG_PID, LOG_DAEMON); + dm_event_log_set(_debug_level, _use_syslog); + dm_log_init(_dmeventd_log); + (void) dm_prepare_selinux_context(DMEVENTD_PIDFILE, S_IFREG); if (dm_create_lockfile(DMEVENTD_PIDFILE) == 0) exit(EXIT_FAILURE);

1 0

master - dmeventd: use log_ macros in plugins
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=c90363b585a41de2… Commit: c90363b585a41de2a65872f9be359db22ee64325 Parent: 915f0faac1fa3541f4a01092536fac5f2ab5982f Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sat Oct 10 16:58:31 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 15:09:03 2015 +0200 dmeventd: use log_ macros in plugins Convert from syslog to standard log_() macros in plugins. --- daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c | 53 +++----------------- daemons/dmeventd/plugins/mirror/dmeventd_mirror.c | 8 +-- daemons/dmeventd/plugins/raid/dmeventd_raid.c | 10 ++-- .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 9 ++-- daemons/dmeventd/plugins/thin/dmeventd_thin.c | 11 ++-- 5 files changed, 23 insertions(+), 68 deletions(-) diff --git a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c index ecaa058..0ba3694 100644 --- a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c +++ b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2010 Red Hat, Inc. All rights reserved. + * Copyright (C) 2010-2015 Red Hat, Inc. All rights reserved. * * This file is part of LVM2. * @@ -13,15 +13,11 @@ */ #include "lib.h" -#include "log.h" - -#include "lvm2cmd.h" #include "dmeventd_lvm.h" +#include "libdevmapper-event.h" +#include "lvm2cmd.h" #include <pthread.h> -#include <syslog.h> - -extern int dmeventd_debug; /* * register_device() is called first and performs initialisation. @@ -36,48 +32,13 @@ static int _register_count = 0; static struct dm_pool *_mem_pool = NULL; static void *_lvm_handle = NULL; +DM_EVENT_LOG_FN("lvm") + /* * Currently only one event can be processed at a time. */ static pthread_mutex_t _event_mutex = PTHREAD_MUTEX_INITIALIZER; -/* - * FIXME Do not pass things directly to syslog, rather use the existing logging - * facilities to sort logging ... however that mechanism needs to be somehow - * configurable and we don't have that option yet - */ -static void _temporary_log_fn(int level, - const char *file __attribute__((unused)), - int line __attribute__((unused)), - int dm_errno __attribute__((unused)), - const char *message) -{ - level &= ~(_LOG_STDERR | _LOG_ONCE); - - switch (level) { - case _LOG_DEBUG: - if (dmeventd_debug >= 3) - syslog(LOG_DEBUG, "%s", message); - break; - case _LOG_INFO: - if (dmeventd_debug >= 2) - syslog(LOG_INFO, "%s", message); - break; - case _LOG_NOTICE: - if (dmeventd_debug >= 1) - syslog(LOG_NOTICE, "%s", message); - break; - case _LOG_WARN: - syslog(LOG_WARNING, "%s", message); - break; - case _LOG_ERR: - syslog(LOG_ERR, "%s", message); - break; - default: - syslog(LOG_CRIT, "%s", message); - } -} - void dmeventd_lvm2_lock(void) { pthread_mutex_lock(&_event_mutex); @@ -95,8 +56,8 @@ int dmeventd_lvm2_init(void) pthread_mutex_lock(&_register_mutex); if (!_lvm_handle) { - if (!getenv("LVM_LOG_FILE_EPOCH")) - lvm2_log_fn(_temporary_log_fn); + lvm2_log_fn((lvm2_log_fn_t)print_log); + if (!(_lvm_handle = lvm2_init())) goto out; diff --git a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c index 288e741..c010794 100644 --- a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c +++ b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2005-2012 Red Hat, Inc. All rights reserved. + * Copyright (C) 2005-2015 Red Hat, Inc. All rights reserved. * * This file is part of LVM2. * @@ -13,20 +13,18 @@ */ #include "lib.h" - #include "libdevmapper-event.h" #include "dmeventd_lvm.h" #include "defaults.h" -#include <syslog.h> /* FIXME Replace syslog with multilog */ -/* FIXME Missing openlog? */ -/* FIXME Replace most syslogs with log_error() style messages and add complete context. */ /* FIXME Reformat to 80 char lines. */ #define ME_IGNORE 0 #define ME_INSYNC 1 #define ME_FAILURE 2 +DM_EVENT_LOG_FN("mirr") + static int _process_status_code(const char status_code, const char *dev_name, const char *dev_type, int r) { diff --git a/daemons/dmeventd/plugins/raid/dmeventd_raid.c b/daemons/dmeventd/plugins/raid/dmeventd_raid.c index 055bfc2..822d7c4 100644 --- a/daemons/dmeventd/plugins/raid/dmeventd_raid.c +++ b/daemons/dmeventd/plugins/raid/dmeventd_raid.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2005-2011 Red Hat, Inc. All rights reserved. + * Copyright (C) 2005-2015 Red Hat, Inc. All rights reserved. * * This file is part of LVM2. * @@ -13,13 +13,11 @@ */ #include "lib.h" - -#include "libdevmapper-event.h" #include "dmeventd_lvm.h" +#include "libdevmapper-event.h" + +DM_EVENT_LOG_FN("raid") -#include <syslog.h> /* FIXME Replace syslog with multilog */ -/* FIXME Missing openlog? */ -/* FIXME Replace most syslogs with log_error() style messages and add complete context. */ /* FIXME Reformat to 80 char lines. */ /* diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index 1372769..f78b8f7 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2007-2011 Red Hat, Inc. All rights reserved. + * Copyright (C) 2007-2015 Red Hat, Inc. All rights reserved. * * This file is part of LVM2. * @@ -13,14 +13,11 @@ */ #include "lib.h" - -#include "libdevmapper-event.h" #include "dmeventd_lvm.h" +#include "libdevmapper-event.h" #include <sys/wait.h> -#include <syslog.h> /* FIXME Replace syslog with multilog */ #include <stdarg.h> -/* FIXME Missing openlog? */ /* First warning when snapshot is 80% full. */ #define WARNING_THRESH 80 @@ -38,6 +35,8 @@ struct dso_state { char cmd_str[1024]; }; +DM_EVENT_LOG_FN("snap") + static int _run(const char *cmd, ...) { va_list ap; diff --git a/daemons/dmeventd/plugins/thin/dmeventd_thin.c b/daemons/dmeventd/plugins/thin/dmeventd_thin.c index febf554..f57eb9b 100644 --- a/daemons/dmeventd/plugins/thin/dmeventd_thin.c +++ b/daemons/dmeventd/plugins/thin/dmeventd_thin.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2011-2013 Red Hat, Inc. All rights reserved. + * Copyright (C) 2011-2015 Red Hat, Inc. All rights reserved. * * This file is part of LVM2. * @@ -12,15 +12,12 @@ * Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#include "lib.h" - -#include "libdevmapper-event.h" +#include "lib.h" /* using here lvm log */ #include "dmeventd_lvm.h" +#include "libdevmapper-event.h" #include <sys/wait.h> -#include <syslog.h> /* FIXME Replace syslog with multilog */ #include <stdarg.h> -/* FIXME Missing openlog? */ /* First warning when thin is 80% full. */ #define WARNING_THRESH 80 @@ -52,6 +49,8 @@ struct dso_state { # define MKDEV(x,y) makedev((x),(y)) #endif +DM_EVENT_LOG_FN("thin") + /* Get dependencies for device, and try to find matching device */ static int _has_deps(const char *name, int tp_major, int tp_minor, int *dev_minor) {

1 0

master - dmeventd: syslog -> log_xxxxx
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=915f0faac1fa3541… Commit: 915f0faac1fa3541f4a01092536fac5f2ab5982f Parent: 0641e3a5fd19628db0e08fca252c13f40230ffa2 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 9 21:57:48 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 14:43:30 2015 +0200 dmeventd: syslog -> log_xxxxx Convert logging of dmeventd to libdm logging macros. --- WHATS_NEW_DM | 1 + daemons/dmeventd/dmeventd.c | 94 +++++++++----------- daemons/dmeventd/libdevmapper-event.c | 34 ++++---- daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c | 6 +- daemons/dmeventd/plugins/mirror/dmeventd_mirror.c | 46 +++++----- daemons/dmeventd/plugins/raid/dmeventd_raid.c | 38 ++++---- .../dmeventd/plugins/snapshot/dmeventd_snapshot.c | 30 ++++--- daemons/dmeventd/plugins/thin/dmeventd_thin.c | 53 ++++++------ 8 files changed, 145 insertions(+), 157 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index a1c121a..ff4ac22 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Convert dmeventd to use common logging macro system from libdm. Return -ENOMEM when device registration fails instead of 0 (=success). Enforce writethrough mode for cleaner policy. Add support for recognition and deactivation of MD devices to blkdeactivate. diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 844b2a6..581ce18 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2005-2007 Red Hat, Inc. All rights reserved. + * Copyright (C) 2005-2015 Red Hat, Inc. All rights reserved. * * This file is part of the device-mapper userspace tools. * @@ -718,13 +718,12 @@ static int _event_wait(struct thread_status *thread, struct dm_task **task) } else if (thread->status == DM_THREAD_SHUTDOWN && ioctl_errno == EINTR) ret = DM_WAIT_FATAL; else { - syslog(LOG_NOTICE, "dm_task_run failed, errno = %d, %s", - ioctl_errno, strerror(ioctl_errno)); if (ioctl_errno == ENXIO) { - syslog(LOG_ERR, "%s disappeared, detaching", - thread->device.name); + log_error("%s disappeared, detaching.", + thread->device.name); ret = DM_WAIT_FATAL; - } + } else + log_sys_error("dm_task_run", ""); } } DEBUGLOG("Completed waitevent task for %s", thread->device.uuid); @@ -778,8 +777,8 @@ static void _monitor_unregister(void *arg) DEBUGLOG("_monitor_unregister thread cleanup handler running"); if (!_do_unregister_device(thread)) - syslog(LOG_ERR, "%s: %s unregister failed\n", __func__, - thread->device.name); + log_error("%s: %s unregister failed.", __func__, + thread->device.name); if (thread->current_task) { dm_task_destroy(thread->current_task); thread->current_task = NULL; @@ -981,8 +980,8 @@ static struct dso_data *_load_dso(struct message_data *data) if (!(dl = dlopen(data->dso_name, RTLD_NOW))) { const char *dlerr = dlerror(); - syslog(LOG_ERR, "dmeventd %s dlopen failed: %s", data->dso_name, - dlerr); + log_error("dmeventd %s dlopen failed: %s.", + data->dso_name, dlerr); data->msg->size = dm_asprintf(&(data->msg->data), "%s %s dlopen failed: %s", data->id, data->dso_name, dlerr); @@ -1309,8 +1308,7 @@ static int _open_fifos(struct dm_event_fifos *fifos) /* Create client fifo. */ (void) dm_prepare_selinux_context(fifos->client_path, S_IFIFO); if ((mkfifo(fifos->client_path, 0600) == -1) && errno != EEXIST) { - syslog(LOG_ERR, "%s: Failed to create client fifo %s: %m.\n", - __func__, fifos->client_path); + log_sys_error("client mkfifo", fifos->client_path); (void) dm_prepare_selinux_context(NULL, 0); goto fail; } @@ -1318,8 +1316,7 @@ static int _open_fifos(struct dm_event_fifos *fifos) /* Create server fifo. */ (void) dm_prepare_selinux_context(fifos->server_path, S_IFIFO); if ((mkfifo(fifos->server_path, 0600) == -1) && errno != EEXIST) { - syslog(LOG_ERR, "%s: Failed to create server fifo %s: %m.\n", - __func__, fifos->server_path); + log_sys_error("server mkfifo", fifos->server_path); (void) dm_prepare_selinux_context(NULL, 0); goto fail; } @@ -1328,58 +1325,53 @@ static int _open_fifos(struct dm_event_fifos *fifos) /* Warn about wrong permissions if applicable */ if ((!stat(fifos->client_path, &st)) && (st.st_mode & 0777) != 0600) - syslog(LOG_WARNING, "Fixing wrong permissions on %s: %m.\n", - fifos->client_path); + log_warn("WARNING: Fixing wrong permissions on %s: %s.\n", + fifos->client_path, strerror(errno)); if ((!stat(fifos->server_path, &st)) && (st.st_mode & 0777) != 0600) - syslog(LOG_WARNING, "Fixing wrong permissions on %s: %m.\n", - fifos->server_path); + log_warn("WARNING: Fixing wrong permissions on %s: %s.\n", + fifos->server_path, strerror(errno)); /* If they were already there, make sure permissions are ok. */ if (chmod(fifos->client_path, 0600)) { - syslog(LOG_ERR, "Unable to set correct file permissions on %s: %m.\n", - fifos->client_path); + log_sys_error("chmod", fifos->client_path); goto fail; } if (chmod(fifos->server_path, 0600)) { - syslog(LOG_ERR, "Unable to set correct file permissions on %s: %m.\n", - fifos->server_path); + log_sys_error("chmod", fifos->server_path); goto fail; } /* Need to open read+write or we will block or fail */ if ((fifos->server = open(fifos->server_path, O_RDWR)) < 0) { - syslog(LOG_ERR, "Failed to open fifo server %s: %m.\n", - fifos->server_path); + log_sys_error("server open", fifos->server_path); goto fail; } if (fcntl(fifos->server, F_SETFD, FD_CLOEXEC) < 0) { - syslog(LOG_ERR, "Failed to set FD_CLOEXEC for fifo server %s: %m.\n", - fifos->server_path); + log_sys_error("fcntl(FD_CLOEXEC)", fifos->server_path); goto fail; } /* Need to open read+write for select() to work. */ if ((fifos->client = open(fifos->client_path, O_RDWR)) < 0) { - syslog(LOG_ERR, "Failed to open fifo client %s: %m", fifos->client_path); + log_sys_error("client open", fifos->client_path); goto fail; } if (fcntl(fifos->client, F_SETFD, FD_CLOEXEC) < 0) { - syslog(LOG_ERR, "Failed to set FD_CLOEXEC for fifo client %s: %m.\n", - fifos->client_path); + log_sys_error("fcntl(FD_CLOEXEC)", fifos->client_path); goto fail; } return 1; fail: if (fifos->server >= 0 && close(fifos->server)) - syslog(LOG_ERR, "Failed to close fifo server %s: %m", fifos->server_path); + log_sys_error("server close", fifos->server_path); if (fifos->client >= 0 && close(fifos->client)) - syslog(LOG_ERR, "Failed to close fifo client %s: %m", fifos->client_path); + log_sys_error("client close", fifos->client_path); return 0; } @@ -1591,7 +1583,7 @@ static void _process_request(struct dm_event_fifos *fifos) if (die) { if (unlink(DMEVENTD_PIDFILE)) - perror(DMEVENTD_PIDFILE ": unlink failed"); + log_sys_error("unlink", DMEVENTD_PIDFILE); _exit(0); } } @@ -1637,15 +1629,14 @@ static void _cleanup_unused_threads(void) if (ret == ESRCH) { thread->status = DM_THREAD_DONE; } else if (ret) { - syslog(LOG_ERR, "Unable to terminate thread: %s", - strerror(ret)); + log_error("Unable to terminate thread: %s", + strerror(ret)); } break; } dm_list_del(l); - syslog(LOG_ERR, - "thread can't be on unused list unless !thread->events"); + log_error("thread can't be on unused list unless !thread->events"); thread->status = DM_THREAD_RUNNING; LINK_THREAD(thread); @@ -1665,7 +1656,7 @@ static void _cleanup_unused_threads(void) _unlock_mutex(); if (join_ret) - syslog(LOG_ERR, "Failed pthread_join: %s\n", strerror(join_ret)); + log_error("Failed pthread_join: %s.", strerror(join_ret)); } static void _sig_alarm(int signum __attribute__((unused))) @@ -1709,14 +1700,14 @@ static int _set_oom_adj(const char *oom_adj_path, int val) FILE *fp; if (!(fp = fopen(oom_adj_path, "w"))) { - perror("oom_adj: fopen failed"); + log_sys_error("open", oom_adj_path); return 0; } fprintf(fp, "%i", val); if (dm_fclose(fp)) - perror("oom_adj: fclose failed"); + log_sys_error("fclose", oom_adj_path); return 1; } @@ -1730,14 +1721,11 @@ static int _protect_against_oom_killer(void) if (stat(OOM_ADJ_FILE, &st) == -1) { if (errno != ENOENT) - perror(OOM_ADJ_FILE ": stat failed"); + log_sys_error("stat", OOM_ADJ_FILE); /* Try old oom_adj interface as a fallback */ if (stat(OOM_ADJ_FILE_OLD, &st) == -1) { - if (errno == ENOENT) - perror(OOM_ADJ_FILE_OLD " not found"); - else - perror(OOM_ADJ_FILE_OLD ": stat failed"); + log_sys_error("stat", OOM_ADJ_FILE_OLD); return 1; } @@ -1826,14 +1814,14 @@ out: static void _remove_files_on_exit(void) { if (unlink(DMEVENTD_PIDFILE)) - perror(DMEVENTD_PIDFILE ": unlink failed"); + log_sys_error("unlink", DMEVENTD_PIDFILE); if (!_systemd_activation) { if (unlink(DM_EVENT_FIFO_CLIENT)) - perror(DM_EVENT_FIFO_CLIENT " : unlink failed"); + log_sys_error("unlink", DM_EVENT_FIFO_CLIENT); if (unlink(DM_EVENT_FIFO_SERVER)) - perror(DM_EVENT_FIFO_SERVER " : unlink failed"); + log_sys_error("unlink", DM_EVENT_FIFO_SERVER); } } @@ -1855,7 +1843,7 @@ static void _daemonize(void) switch (pid = fork()) { case -1: - perror("fork failed:"); + log_sys_error("fork", ""); exit(EXIT_FAILURE); case 0: /* Child */ @@ -2167,7 +2155,7 @@ int main(int argc, char *argv[]) #ifdef __linux__ /* Systemd has adjusted oom killer for us already */ if (!_systemd_activation && !_protect_against_oom_killer()) - syslog(LOG_ERR, "Failed to protect against OOM killer"); + log_error("Failed to protect against OOM killer."); #endif _init_thread_signals(); @@ -2185,7 +2173,7 @@ int main(int argc, char *argv[]) /* Signal parent, letting them know we are ready to go. */ if (!_foreground) kill(getppid(), SIGTERM); - syslog(LOG_NOTICE, "dmeventd ready for processing."); + log_notice("dmeventd ready for processing."); if (_initial_registrations) _process_initial_registrations(); @@ -2204,8 +2192,8 @@ int main(int argc, char *argv[]) _unlock_mutex(); if (nothreads) break; - syslog(LOG_ERR, "There are still devices being monitored."); - syslog(LOG_ERR, "Refusing to exit."); + log_error("There are still devices being monitored."); + log_error("Refusing to exit."); } _process_request(&fifos); _cleanup_unused_threads(); @@ -2215,7 +2203,7 @@ int main(int argc, char *argv[]) pthread_mutex_destroy(&_global_mutex); - syslog(LOG_NOTICE, "dmeventd shutting down."); + log_notice("dmeventd shutting down."); closelog(); exit(EXIT_SUCCESS); diff --git a/daemons/dmeventd/libdevmapper-event.c b/daemons/dmeventd/libdevmapper-event.c index 9b04b02..ee4a35c 100644 --- a/daemons/dmeventd/libdevmapper-event.c +++ b/daemons/dmeventd/libdevmapper-event.c @@ -198,7 +198,7 @@ static int _check_message_id(struct dm_event_daemon_message *msg) if ((sscanf(msg->data, "%d:%d", &pid, &seq_nr) != 2) || (pid != getpid()) || (seq_nr != _sequence_nr)) { log_error("Ignoring out-of-sequence reply from dmeventd. " - "Expected %d:%d but received %s", getpid(), + "Expected %d:%d but received %s.", getpid(), _sequence_nr, msg->data); return 0; } @@ -233,7 +233,7 @@ static int _daemon_read(struct dm_event_fifos *fifos, FD_SET(fifos->server, &fds); ret = select(fifos->server + 1, &fds, NULL, NULL, &tval); if (ret < 0 && errno != EINTR) { - log_error("Unable to read from event server"); + log_error("Unable to read from event server."); return 0; } if ((ret == 0) && (i > 4) && !bytes) { @@ -299,7 +299,7 @@ static int _daemon_write(struct dm_event_fifos *fifos, if (ret < 0) { if (errno == EINTR) continue; - log_error("Unable to talk to event daemon"); + log_error("Unable to talk to event daemon."); return 0; } if (ret == 0) @@ -308,7 +308,7 @@ static int _daemon_write(struct dm_event_fifos *fifos, if (ret < 0) { if ((errno == EINTR) || (errno == EAGAIN)) continue; - log_error("Unable to talk to event daemon"); + log_error("Unable to talk to event daemon."); return 0; } } @@ -320,7 +320,7 @@ static int _daemon_write(struct dm_event_fifos *fifos, FD_SET(fifos->client, &fds); ret = select(fifos->client + 1, NULL, &fds, NULL, NULL); if ((ret < 0) && (errno != EINTR)) { - log_error("Unable to talk to event daemon"); + log_error("Unable to talk to event daemon."); return 0; } } while (ret < 1); @@ -330,7 +330,7 @@ static int _daemon_write(struct dm_event_fifos *fifos, if ((errno == EINTR) || (errno == EAGAIN)) continue; else { - log_error("Unable to talk to event daemon"); + log_error("Unable to talk to event daemon."); return 0; } } @@ -360,7 +360,7 @@ int daemon_talk(struct dm_event_fifos *fifos, getpid(), _sequence_nr, dso_name ? : "-", dev_name ? : "-", evmask, timeout))) < 0) { - log_error("_daemon_talk: message allocation failed"); + log_error("_daemon_talk: message allocation failed."); return -ENOMEM; } msg->cmd = cmd; @@ -448,11 +448,11 @@ static int _start_daemon(char *dmeventd_path, struct dm_event_fifos *fifos) else if (!pid) { execvp(args[0], args); - log_error("Unable to exec dmeventd: %s", strerror(errno)); + log_error("Unable to exec dmeventd: %s.", strerror(errno)); _exit(EXIT_FAILURE); } else { if (waitpid(pid, &status, 0) < 0) - log_error("Unable to start dmeventd: %s", + log_error("Unable to start dmeventd: %s.", strerror(errno)); else if (WEXITSTATUS(status)) log_error("Unable to start dmeventd."); @@ -525,7 +525,7 @@ static struct dm_task *_get_device_info(const struct dm_event_handler *dmevh) struct dm_info info; if (!(dmt = dm_task_create(DM_DEVICE_INFO))) { - log_error("_get_device_info: dm_task creation for info failed"); + log_error("_get_device_info: dm_task creation for info failed."); return NULL; } @@ -543,17 +543,17 @@ static struct dm_task *_get_device_info(const struct dm_event_handler *dmevh) /* FIXME Add name or uuid or devno to messages */ if (!dm_task_run(dmt)) { - log_error("_get_device_info: dm_task_run() failed"); + log_error("_get_device_info: dm_task_run() failed."); goto bad; } if (!dm_task_get_info(dmt, &info)) { - log_error("_get_device_info: failed to get info for device"); + log_error("_get_device_info: failed to get info for device."); goto bad; } if (!info.exists) { - log_error("_get_device_info: %s%s%s%.0d%s%.0d%s%s: device not found", + log_error("_get_device_info: %s%s%s%.0d%s%.0d%s%s: device not found.", dmevh->uuid ? : "", (!dmevh->uuid && dmevh->dev_name) ? dmevh->dev_name : "", (!dmevh->uuid && !dmevh->dev_name && dmevh->major > 0) ? "(" : "", @@ -622,12 +622,12 @@ int dm_event_register_handler(const struct dm_event_handler *dmevh) !strstr(dmevh->dso, "libdevmapper-event-lvm2snapshot.so") && !strstr(dmevh->dso, "libdevmapper-event-lvm2mirror.so") && !strstr(dmevh->dso, "libdevmapper-event-lvm2raid.so")) - log_warn("WARNING: %s: dmeventd plugins are deprecated", dmevh->dso); + log_warn("WARNING: %s: dmeventd plugins are deprecated.", dmevh->dso); if ((err = _do_event(DM_EVENT_CMD_REGISTER_FOR_EVENT, dmevh->dmeventd_path, &msg, dmevh->dso, uuid, dmevh->mask, dmevh->timeout)) < 0) { - log_error("%s: event registration failed: %s", + log_error("%s: event registration failed: %s.", dm_task_get_name(dmt), msg.data ? msg.data : strerror(-err)); ret = 0; @@ -654,7 +654,7 @@ int dm_event_unregister_handler(const struct dm_event_handler *dmevh) if ((err = _do_event(DM_EVENT_CMD_UNREGISTER_FOR_EVENT, dmevh->dmeventd_path, &msg, dmevh->dso, uuid, dmevh->mask, dmevh->timeout)) < 0) { - log_error("%s: event deregistration failed: %s", + log_error("%s: event deregistration failed: %s.", dm_task_get_name(dmt), msg.data ? msg.data : strerror(-err)); ret = 0; @@ -933,7 +933,7 @@ int dm_event_get_timeout(const char *device_path, uint32_t *timeout) 0, 0))) { char *p = _skip_string(msg.data, ' '); if (!p) { - log_error("malformed reply from dmeventd '%s'\n", + log_error("Malformed reply from dmeventd '%s'.", msg.data); dm_free(msg.data); return -EIO; diff --git a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c index 7350623..ecaa058 100644 --- a/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c +++ b/daemons/dmeventd/plugins/lvm2/dmeventd_lvm.c @@ -155,8 +155,8 @@ int dmeventd_lvm2_command(struct dm_pool *mem, char *buffer, size_t size, int r; if (!dm_split_lvm_name(mem, device, &vg, &lv, &layer)) { - syslog(LOG_ERR, "Unable to determine VG name from %s.\n", - device); + log_error("Unable to determine VG name from %s.", + device); return 0; } @@ -170,7 +170,7 @@ int dmeventd_lvm2_command(struct dm_pool *mem, char *buffer, size_t size, dm_pool_free(mem, vg); if (r < 0) { - syslog(LOG_ERR, "Unable to form LVM command. (too long).\n"); + log_error("Unable to form LVM command. (too long)."); return 0; } diff --git a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c index 084be28..288e741 100644 --- a/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c +++ b/daemons/dmeventd/plugins/mirror/dmeventd_mirror.c @@ -39,18 +39,15 @@ static int _process_status_code(const char status_code, const char *dev_name, * U => Unclassified failure (bug) */ if (status_code == 'F') { - syslog(LOG_ERR, "%s device %s flush failed.", - dev_type, dev_name); + log_error("%s device %s flush failed.", dev_type, dev_name); r = ME_FAILURE; } else if (status_code == 'S') - syslog(LOG_ERR, "%s device %s sync failed.", - dev_type, dev_name); + log_error("%s device %s sync failed.", dev_type, dev_name); else if (status_code == 'R') - syslog(LOG_ERR, "%s device %s read failed.", - dev_type, dev_name); + log_error("%s device %s read failed.", dev_type, dev_name); else if (status_code != 'A') { - syslog(LOG_ERR, "%s device %s has failed (%c).", - dev_type, dev_name, status_code); + log_error("%s device %s has failed (%c).", + dev_type, dev_name, status_code); r = ME_FAILURE; } @@ -125,7 +122,8 @@ out: out_parse: dm_free(args); - syslog(LOG_ERR, "Unable to parse mirror status string."); + log_error("Unable to parse mirror status string."); + return ME_IGNORE; } @@ -142,7 +140,7 @@ static int _remove_failed_devices(const char *device) r = dmeventd_lvm2_run(cmd_str); if (!r) - syslog(LOG_INFO, "Re-scan of mirror device %s failed.", device); + log_info("Re-scan of mirror device %s failed.", device); if (!dmeventd_lvm2_command(dmeventd_lvm2_pool(), cmd_str, sizeof(cmd_str), "lvconvert --config devices{ignore_suspended_devices=1} " @@ -152,8 +150,8 @@ static int _remove_failed_devices(const char *device) /* if repair goes OK, report success even if lvscan has failed */ r = dmeventd_lvm2_run(cmd_str); - syslog(LOG_INFO, "Repair of mirrored device %s %s.", device, - (r) ? "finished successfully" : "failed"); + log_info("Repair of mirrored device %s %s.", device, + (r) ? "finished successfully" : "failed"); return (r) ? 0 : -1; } @@ -175,12 +173,12 @@ void process_event(struct dm_task *dmt, &target_type, &params); if (!target_type) { - syslog(LOG_INFO, "%s mapping lost.", device); + log_info("%s mapping lost.", device); continue; } if (strcmp(target_type, "mirror")) { - syslog(LOG_INFO, "%s has unmirrored portion.", device); + log_info("%s has unmirrored portion.", device); continue; } @@ -190,25 +188,25 @@ void process_event(struct dm_task *dmt, _part_ of the device is in sync Also, this is not an error */ - syslog(LOG_NOTICE, "%s is now in-sync.", device); + log_notice("%s is now in-sync.", device); break; case ME_FAILURE: - syslog(LOG_ERR, "Device failure in %s.", device); + log_error("Device failure in %s.", device); if (_remove_failed_devices(device)) /* FIXME Why are all the error return codes unused? Get rid of them? */ - syslog(LOG_ERR, "Failed to remove faulty devices in %s.", - device); + log_error("Failed to remove faulty devices in %s.", + device); /* Should check before warning user that device is now linear else - syslog(LOG_NOTICE, "%s is now a linear device.\n", - device); + log_notice("%s is now a linear device.", + device); */ break; case ME_IGNORE: break; default: /* FIXME Provide value then! */ - syslog(LOG_INFO, "Unknown event received."); + log_info("Unknown event received."); } } while (next); @@ -224,7 +222,7 @@ int register_device(const char *device, if (!dmeventd_lvm2_init()) return 0; - syslog(LOG_INFO, "Monitoring mirror device %s for events.", device); + log_info("Monitoring mirror device %s for events.", device); return 1; } @@ -235,8 +233,8 @@ int unregister_device(const char *device, int minor __attribute__((unused)), void **unused __attribute__((unused))) { - syslog(LOG_INFO, "No longer monitoring mirror device %s for events.", - device); + log_info("No longer monitoring mirror device %s for events.", + device); dmeventd_lvm2_exit(); return 1; diff --git a/daemons/dmeventd/plugins/raid/dmeventd_raid.c b/daemons/dmeventd/plugins/raid/dmeventd_raid.c index 4690cfa..055bfc2 100644 --- a/daemons/dmeventd/plugins/raid/dmeventd_raid.c +++ b/daemons/dmeventd/plugins/raid/dmeventd_raid.c @@ -39,7 +39,7 @@ static int run_repair(const char *device) r = dmeventd_lvm2_run(cmd_str); if (!r) - syslog(LOG_INFO, "Re-scan of RAID device %s failed.", device); + log_info("Re-scan of RAID device %s failed.", device); if (!dmeventd_lvm2_command(dmeventd_lvm2_pool(), cmd_str, sizeof(cmd_str), "lvconvert --config devices{ignore_suspended_devices=1} " @@ -50,7 +50,7 @@ static int run_repair(const char *device) r = dmeventd_lvm2_run(cmd_str); if (!r) - syslog(LOG_INFO, "Repair of RAID device %s failed.", device); + log_info("Repair of RAID device %s failed.", device); return (r) ? 0 : -1; } @@ -69,8 +69,7 @@ static int _process_raid_event(char *params, const char *device) * <health chars> <resync ratio> */ if (!dm_split_words(params, 4, 0, a)) { - syslog(LOG_ERR, "Failed to process status line for %s\n", - device); + log_error("Failed to process status line for %s.", device); return -EINVAL; } raid_type = a[0]; @@ -79,8 +78,8 @@ static int _process_raid_event(char *params, const char *device) resync_ratio = a[3]; if (!(n = atoi(num_devices))) { - syslog(LOG_ERR, "Failed to parse number of devices for %s: %s", - device, num_devices); + log_error("Failed to parse number of devices for %s: %s.", + device, num_devices); return -EINVAL; } @@ -92,9 +91,8 @@ static int _process_raid_event(char *params, const char *device) /* Device is 'a'live, but not yet in-sync */ break; case 'D': - syslog(LOG_ERR, - "Device #%d of %s array, %s, has failed.", - i, raid_type, device); + log_error("Device #%d of %s array, %s, has failed.", + i, raid_type, device); failure++; break; default: @@ -107,13 +105,13 @@ static int _process_raid_event(char *params, const char *device) p = strstr(resync_ratio, "/"); if (!p) { - syslog(LOG_ERR, "Failed to parse resync_ratio for %s: %s", - device, resync_ratio); + log_error("Failed to parse resync_ratio for %s: %s.", + device, resync_ratio); return -EINVAL; } p[0] = '\0'; - syslog(LOG_INFO, "%s array, %s, is %s in-sync.", - raid_type, device, strcmp(resync_ratio, p+1) ? "not" : "now"); + log_info("%s array, %s, is %s in-sync.", + raid_type, device, strcmp(resync_ratio, p+1) ? "not" : "now"); return 0; } @@ -135,18 +133,18 @@ void process_event(struct dm_task *dmt, &target_type, &params); if (!target_type) { - syslog(LOG_INFO, "%s mapping lost.", device); + log_info("%s mapping lost.", device); continue; } if (strcmp(target_type, "raid")) { - syslog(LOG_INFO, "%s has non-raid portion.", device); + log_info("%s has non-raid portion.", device); continue; } if (_process_raid_event(params, device)) - syslog(LOG_ERR, "Failed to process event for %s", - device); + log_error("Failed to process event for %s.", + device); } while (next); dmeventd_lvm2_unlock(); @@ -161,7 +159,7 @@ int register_device(const char *device, if (!dmeventd_lvm2_init()) return 0; - syslog(LOG_INFO, "Monitoring RAID device %s for events.", device); + log_info("Monitoring RAID device %s for events.", device); return 1; } @@ -172,8 +170,8 @@ int unregister_device(const char *device, int minor __attribute__((unused)), void **unused __attribute__((unused))) { - syslog(LOG_INFO, "No longer monitoring RAID device %s for events.", - device); + log_info("No longer monitoring RAID device %s for events.", + device); dmeventd_lvm2_exit(); return 1; diff --git a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c index 1b1d977..1372769 100644 --- a/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c +++ b/daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c @@ -62,7 +62,7 @@ static int _run(const char *cmd, ...) va_end(ap); execvp(cmd, (char **)argv); - syslog(LOG_ERR, "Failed to execute %s: %s.\n", cmd, strerror(errno)); + log_sys_error("exec", cmd); exit(127); } @@ -90,9 +90,11 @@ static void _umount(const char *device, int major, int minor) char buffer[4096]; char *words[3]; struct stat st; + const char procmounts[] = "/proc/mounts"; - if (!(mounts = fopen("/proc/mounts", "r"))) { - syslog(LOG_ERR, "Could not read /proc/mounts. Not umounting %s.\n", device); + if (!(mounts = fopen(procmounts, "r"))) { + log_sys_error("fopen", procmounts); + log_error("Not umounting %s.", device); return; } @@ -112,15 +114,15 @@ static void _umount(const char *device, int major, int minor) if (S_ISBLK(st.st_mode) && major(st.st_rdev) == major && minor(st.st_rdev) == minor) { - syslog(LOG_ERR, "Unmounting invalid snapshot %s from %s.\n", device, words[1]); - if (!_run(UMOUNT_COMMAND, "-fl", words[1], NULL)) - syslog(LOG_ERR, "Failed to umount snapshot %s from %s: %s.\n", - device, words[1], strerror(errno)); + log_error("Unmounting invalid snapshot %s from %s.", device, words[1]); + if (!_run(UMOUNT_COMMAND, "-fl", words[1], NULL)) + log_error("Failed to umount snapshot %s from %s: %s.", + device, words[1], strerror(errno)); } } if (fclose(mounts)) - syslog(LOG_ERR, "Failed to close /proc/mounts.\n"); + log_sys_error("close", procmounts); } void process_event(struct dm_task *dmt, @@ -169,7 +171,7 @@ void process_event(struct dm_task *dmt, * the status string. Report the full status string to syslog. */ if (status->invalid || !status->total_sectors) { - syslog(LOG_ERR, "Snapshot %s changed state to: %s\n", device, params); + log_error("Snapshot %s changed state to: %s.", device, params); state->percent_check = 0; goto out; } @@ -181,10 +183,10 @@ void process_event(struct dm_task *dmt, state->percent_check = (percent / CHECK_STEP) * CHECK_STEP + CHECK_STEP; if (percent >= WARNING_THRESH) /* Print a warning to syslog. */ - syslog(LOG_WARNING, "Snapshot %s is now %i%% full.\n", device, percent); + log_warn("WARNING: Snapshot %s is now %i%% full.", device, percent); /* Try to extend the snapshot, in accord with user-set policies */ if (!_extend(state->cmd_str)) - syslog(LOG_ERR, "Failed to extend snapshot %s.\n", device); + log_error("Failed to extend snapshot %s.", device); } out: @@ -218,7 +220,7 @@ int register_device(const char *device, state->percent_check = CHECK_MINIMUM; *private = state; - syslog(LOG_INFO, "Monitoring snapshot %s\n", device); + log_info("Monitoring snapshot %s.", device); return 1; bad: @@ -226,7 +228,7 @@ bad: dm_pool_destroy(statemem); dmeventd_lvm2_exit(); out: - syslog(LOG_ERR, "Failed to monitor snapshot %s.\n", device); + log_error("Failed to monitor snapshot %s.", device); return 0; } @@ -239,7 +241,7 @@ int unregister_device(const char *device, { struct dso_state *state = *private; - syslog(LOG_INFO, "No longer monitoring snapshot %s\n", device); + log_info("No longer monitoring snapshot %s.", device); dm_pool_destroy(state->mem); dmeventd_lvm2_exit(); diff --git a/daemons/dmeventd/plugins/thin/dmeventd_thin.c b/daemons/dmeventd/plugins/thin/dmeventd_thin.c index 3f057a6..febf554 100644 --- a/daemons/dmeventd/plugins/thin/dmeventd_thin.c +++ b/daemons/dmeventd/plugins/thin/dmeventd_thin.c @@ -93,8 +93,8 @@ static int _has_deps(const char *name, int tp_major, int tp_minor, int *dev_mino { char dev_name[PATH_MAX]; if (dm_device_get_name(major, minor, 0, dev_name, sizeof(dev_name))) - syslog(LOG_DEBUG, "Found %s (%u:%u) depends on %s", - name, major, *dev_minor, dev_name); + log_debug("Found %s (%u:%u) depends on %s.", + name, major, *dev_minor, dev_name); } #endif r = 1; @@ -144,7 +144,7 @@ out: static int _extend(struct dso_state *state) { #if THIN_DEBUG - syslog(LOG_INFO, "dmeventd executes: %s.\n", state->cmd_str); + log_info("dmeventd executes: %s.", state->cmd_str); #endif return dmeventd_lvm2_run(state->cmd_str); } @@ -173,7 +173,7 @@ static int _run(const char *cmd, ...) va_end(ap); execvp(cmd, (char **)argv); - syslog(LOG_ERR, "Failed to execute %s: %s.\n", cmd, strerror(errno)); + log_sys_error("exec", cmd); exit(127); } @@ -202,11 +202,11 @@ static int _umount_device(char *buffer, unsigned major, unsigned minor, struct mountinfo_s *data = cb_data; if ((major == data->info.major) && dm_bit(data->minors, minor)) { - syslog(LOG_INFO, "Unmounting thin volume %s from %s.\n", - data->device, target); + log_info("Unmounting thin volume %s from %s.", + data->device, target); if (!_run(UMOUNT_COMMAND, "-fl", target, NULL)) - syslog(LOG_ERR, "Failed to umount thin %s from %s: %s.\n", - data->device, target, strerror(errno)); + log_error("Failed to umount thin %s from %s: %s.", + data->device, target, strerror(errno)); } return 1; @@ -230,17 +230,17 @@ static void _umount(struct dm_task *dmt, const char *device) dmeventd_lvm2_unlock(); if (!(data.minors = dm_bitset_create(NULL, MINORS))) { - syslog(LOG_ERR, "Failed to allocate bitset. Not unmounting %s.\n", device); + log_error("Failed to allocate bitset. Not unmounting %s.", device); goto out; } if (!_find_all_devs(data.minors, data.info.major, data.info.minor)) { - syslog(LOG_ERR, "Failed to detect mounted volumes for %s.\n", device); + log_error("Failed to detect mounted volumes for %s.", device); goto out; } if (!dm_mountinfo_read(_umount_device, &data)) { - syslog(LOG_ERR, "Could not parse mountinfo file.\n"); + log_error("Could not parse mountinfo file."); goto out; } @@ -273,21 +273,21 @@ void process_event(struct dm_task *dmt, dm_get_next_target(dmt, next, &start, &length, &target_type, &params); if (!target_type || (strcmp(target_type, "thin-pool") != 0)) { - syslog(LOG_ERR, "Invalid target type.\n"); + log_error("Invalid target type."); goto out; } if (!dm_get_status_thin_pool(state->mem, params, &tps)) { - syslog(LOG_ERR, "Failed to parse status.\n"); + log_error("Failed to parse status."); _umount(dmt, device); goto out; } #if THIN_DEBUG - syslog(LOG_INFO, "%p: Got status %" PRIu64 " / %" PRIu64 - " %" PRIu64 " / %" PRIu64 ".\n", state, - tps->used_metadata_blocks, tps->total_metadata_blocks, - tps->used_data_blocks, tps->total_data_blocks); + log_debug("%p: Got status " FMTu64 " / " FMTu64 " " FMTu64 + " / " FMTu64 ".", state, + tps->used_metadata_blocks, tps->total_metadata_blocks, + tps->used_data_blocks, tps->total_data_blocks); #endif /* Thin pool size had changed. Clear the threshold. */ @@ -311,12 +311,12 @@ void process_event(struct dm_task *dmt, /* FIXME: extension of metadata needs to be written! */ if (percent >= WARNING_THRESH) /* Print a warning to syslog. */ - syslog(LOG_WARNING, "Thin metadata %s is now %i%% full.\n", - device, percent); + log_warn("WARNING: Thin metadata %s is now %i%% full.", + device, percent); /* Try to extend the metadata, in accord with user-set policies */ if (!_extend(state)) { - syslog(LOG_ERR, "Failed to extend thin metadata %s.\n", - device); + log_error("Failed to extend thin metadata %s.", + device); _umount(dmt, device); } /* FIXME: hmm READ-ONLY switch should happen in error path */ @@ -331,10 +331,11 @@ void process_event(struct dm_task *dmt, state->data_percent_check = (percent / CHECK_STEP) * CHECK_STEP + CHECK_STEP; if (percent >= WARNING_THRESH) /* Print a warning to syslog. */ - syslog(LOG_WARNING, "Thin %s is now %i%% full.\n", device, percent); + log_warn("WARNING: Thin %s is now %i%% full.", + device, percent); /* Try to extend the thin data, in accord with user-set policies */ if (!_extend(state)) { - syslog(LOG_ERR, "Failed to extend thin %s.\n", device); + log_error("Failed to extend thin %s.", device); state->data_percent_check = 0; _umount(dmt, device); } @@ -376,11 +377,11 @@ int register_device(const char *device, state->data_percent_check = CHECK_MINIMUM; *private = state; - syslog(LOG_INFO, "Monitoring thin %s.\n", device); + log_info("Monitoring thin %s.", device); return 1; bad: - syslog(LOG_ERR, "Failed to monitor thin %s.\n", device); + log_error("Failed to monitor thin %s.", device); return 0; } @@ -393,7 +394,7 @@ int unregister_device(const char *device, { struct dso_state *state = *private; - syslog(LOG_INFO, "No longer monitoring thin %s.\n", device); + log_info("No longer monitoring thin %s.", device); dm_pool_destroy(state->mem); dmeventd_lvm2_exit();

1 0

master - libdmevent: add dm_event_log
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=0641e3a5fd19628d… Commit: 0641e3a5fd19628db0e08fca252c13f40230ffa2 Parent: 11a084cf42e79f130e4389bac57688cfd6657fc1 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sat Oct 10 16:40:52 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 14:40:46 2015 +0200 libdmevent: add dm_event_log Add dm_event_log() logging function. Controled by dm_event_log_set(). --- daemons/dmeventd/libdevmapper-event.c | 81 ++++++++++++++++++++++++++++++++- daemons/dmeventd/libdevmapper-event.h | 21 ++++++++- 2 files changed, 99 insertions(+), 3 deletions(-) diff --git a/daemons/dmeventd/libdevmapper-event.c b/daemons/dmeventd/libdevmapper-event.c index 81f18c9..9b04b02 100644 --- a/daemons/dmeventd/libdevmapper-event.c +++ b/daemons/dmeventd/libdevmapper-event.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2005-2007 Red Hat, Inc. All rights reserved. + * Copyright (C) 2005-2015 Red Hat, Inc. All rights reserved. * * This file is part of the device-mapper userspace tools. * @@ -12,9 +12,9 @@ * Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#include "dm-logging.h" #include "dmlib.h" #include "libdevmapper-event.h" -//#include "libmultilog.h" #include "dmeventd.h" #include <fcntl.h> @@ -23,7 +23,11 @@ #include <sys/stat.h> #include <sys/wait.h> #include <arpa/inet.h> /* for htonl, ntohl */ +#include <pthread.h> +#include <syslog.h> +static int _debug_level = 0; +static int _use_syslog = 0; static int _sequence_nr = 0; struct dm_event_handler { @@ -823,6 +827,79 @@ int dm_event_get_version(struct dm_event_fifos *fifos, int *version) { return 1; } +void dm_event_log_set(int debug_level, int use_syslog) +{ + _debug_level = debug_level; + _use_syslog = use_syslog; +} + +void dm_event_log(const char *subsys, int level, const char *file, + int line, int dm_errno_or_class, + const char *format, va_list ap) +{ + static pthread_mutex_t _log_mutex = PTHREAD_MUTEX_INITIALIZER; + static time_t start = 0; + const char *indent = ""; + FILE *stream = stdout; + int prio = -1; + time_t now; + + switch (level & ~(_LOG_STDERR | _LOG_ONCE)) { + case _LOG_DEBUG: + if (_debug_level < 3) + return; + prio = LOG_DEBUG; + indent = " "; + break; + case _LOG_INFO: + if (_debug_level < 2) + return; + prio = LOG_INFO; + indent = " "; + break; + case _LOG_NOTICE: + if (_debug_level < 1) + return; + prio = LOG_NOTICE; + indent = " "; + break; + case _LOG_WARN: + prio = LOG_WARNING; + break; + case _LOG_ERR: + prio = LOG_ERR; + stream = stderr; + break; + default: + prio = LOG_CRIT; + } + + /* Serialize to keep lines readable */ + pthread_mutex_lock(&_log_mutex); + + if (_use_syslog) { + vsyslog(prio, format, ap); + } else { + now = time(NULL); + if (!start) + start = now; + now -= start; + fprintf(stream, "[%2d:%02d] %8x:%-6s%s", + (int)now / 60, (int)now % 60, + // TODO: Maybe use shorter ID + // ((int)(pthread_self()) >> 6) & 0xffff, + (int)pthread_self(), subsys, + (_debug_level > 3) ? "" : indent); + if (_debug_level > 3) + fprintf(stream, "%28s:%4d %s", file, line, indent); + vfprintf(stream, _(format), ap); + fputc('\n', stream); + fflush(stream); + } + + pthread_mutex_unlock(&_log_mutex); +} + #if 0 /* left out for now */ static char *_skip_string(char *src, const int delimiter) diff --git a/daemons/dmeventd/libdevmapper-event.h b/daemons/dmeventd/libdevmapper-event.h index 532bebf..a00702a 100644 --- a/daemons/dmeventd/libdevmapper-event.h +++ b/daemons/dmeventd/libdevmapper-event.h @@ -1,5 +1,5 @@ /* - * Copyright (C) 2005-2007 Red Hat, Inc. All rights reserved. + * Copyright (C) 2005-2015 Red Hat, Inc. All rights reserved. * * This file is part of the device-mapper userspace tools. * @@ -105,6 +105,25 @@ int dm_event_get_registered_device(struct dm_event_handler *dmevh, int next); int dm_event_register_handler(const struct dm_event_handler *dmevh); int dm_event_unregister_handler(const struct dm_event_handler *dmevh); +/* Set debug level for logging, and whether to log on stdout/stderr or syslog */ +void dm_event_log_set(int debug_level, int use_syslog); + +/* Log messages acroding to current debug level */ +__attribute__((format(printf, 6, 0))) +void dm_event_log(const char *subsys, int level, const char *file, + int line, int dm_errno_or_class, + const char *format, va_list ap); +/* Macro to route print_log do dm_event_log() */ +#define DM_EVENT_LOG_FN(subsys) \ +void print_log(int level, const char *file, int line, int dm_errno_or_class,\ + const char *format, ...)\ +{\ + va_list ap;\ + va_start(ap, format);\ + dm_event_log(subsys, level, file, line, dm_errno_or_class, format, ap);\ + va_end(ap);\ +} + /* Prototypes for DSO interface, see dmeventd.c, struct dso_data for detailed descriptions. */ // FIXME misuse of bitmask as enum

1 0

master - dmeventd: log all commands with CMD: prefix
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=11a084cf42e79f13… Commit: 11a084cf42e79f130e4389bac57688cfd6657fc1 Parent: d60794c3a3948f5ec5d4c529569ba66d9a635e55 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 11:19:59 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 14:40:46 2015 +0200 dmeventd: log all commands with CMD: prefix For easier grep of all processed command - add CMD: prefix to them. Use >>> for entring command processing and <<< when command is processed. --- daemons/dmeventd/dmeventd.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 37cf8e9..844b2a6 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -1574,7 +1574,7 @@ static void _process_request(struct dm_event_fifos *fifos) return; cmd = msg.cmd; - DEBUGLOG("%s (0x%x) processing...", decode_cmd(cmd), cmd); + DEBUGLOG(">>> CMD:%s (0x%x) processing...", decode_cmd(cmd), cmd); die = (msg.cmd == DM_EVENT_CMD_DIE) ? 1 : 0; @@ -1587,7 +1587,7 @@ static void _process_request(struct dm_event_fifos *fifos) dm_free(msg.data); - DEBUGLOG("%s (0x%x) completed (=%d).", decode_cmd(cmd), cmd, msg.cmd); + DEBUGLOG("<<< CMD:%s (0x%x) completed (=%d).", decode_cmd(cmd), cmd, msg.cmd); if (die) { if (unlink(DMEVENTD_PIDFILE))

1 0

master - dmeventd: fix loging report
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=d60794c3a3948f5e… Commit: d60794c3a3948f5ec5d4c529569ba66d9a635e55 Parent: 1b1c01a27b359f8e91c3c9e08684c435d8fd51c2 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 9 22:16:33 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 14:40:46 2015 +0200 dmeventd: fix loging report msg.cmd holds return value from processing after _do_process_request, so save it before call. --- daemons/dmeventd/dmeventd.c | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index ffc941f..37cf8e9 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -1564,6 +1564,7 @@ static void _process_request(struct dm_event_fifos *fifos) { int die; struct dm_event_daemon_message msg = { 0 }; + int cmd; /* * Read the request from the client (client_read, client_write @@ -1572,7 +1573,8 @@ static void _process_request(struct dm_event_fifos *fifos) if (!_client_read(fifos, &msg)) return; - DEBUGLOG("%s (0x%x) processing...", decode_cmd(msg.cmd), msg.cmd); + cmd = msg.cmd; + DEBUGLOG("%s (0x%x) processing...", decode_cmd(cmd), cmd); die = (msg.cmd == DM_EVENT_CMD_DIE) ? 1 : 0; @@ -1585,7 +1587,7 @@ static void _process_request(struct dm_event_fifos *fifos) dm_free(msg.data); - DEBUGLOG("%s (0x%x) completed.", decode_cmd(msg.cmd), msg.cmd); + DEBUGLOG("%s (0x%x) completed (=%d).", decode_cmd(cmd), cmd, msg.cmd); if (die) { if (unlink(DMEVENTD_PIDFILE))

1 0

master - log: no file for external logging
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=1b1c01a27b359f8e… Commit: 1b1c01a27b359f8e91c3c9e08684c435d8fd51c2 Parent: 72d700b064fb64f6bb7e20c23dbb2ce8b6d8feb7 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 09:44:33 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 14:40:44 2015 +0200 log: no file for external logging When external logging is set do not bother creating log file, as it's not going to be used. --- WHATS_NEW | 1 + lib/log/log.c | 4 ++++ 2 files changed, 5 insertions(+), 0 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index c089228..7507b9c 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Avoid creation of log file when logging function is set. Do not change logging in lvm2 library when it's already set. Check for enough space in thin-pool in command before creating new thin. Make libblkid detect all copies of the same signature if use_blkid_wiping=1. diff --git a/lib/log/log.c b/lib/log/log.c index bd11363..7f31361 100644 --- a/lib/log/log.c +++ b/lib/log/log.c @@ -74,6 +74,10 @@ void init_log_file(const char *log_file, int append) int i = 0; _log_file_path[0] = '\0'; + + if (_lvm2_log_fn) + return; /* No log file, when log function is set */ + if ((env = getenv("LVM_LOG_FILE_EPOCH"))) { while (isalpha(env[i]) && i < 32) /* Up to 32 alphas */ i++;

1 0

master - toolcontext: do not change already set dm_log
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=72d700b064fb64f6… Commit: 72d700b064fb64f6bb7e20c23dbb2ce8b6d8feb7 Parent: 86b04ebd193c9db09ee5f03a2de4997aa2ab52c3 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 9 21:38:50 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 14:39:37 2015 +0200 toolcontext: do not change already set dm_log If the dm_log is already set to non-default logger, avoid to change it. --- WHATS_NEW | 1 + lib/commands/toolcontext.c | 3 ++- 2 files changed, 3 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 348c099..c089228 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Do not change logging in lvm2 library when it's already set. Check for enough space in thin-pool in command before creating new thin. Make libblkid detect all copies of the same signature if use_blkid_wiping=1. Fix vgimportclone with -n to not add number unnecessarily to base VG name. diff --git a/lib/commands/toolcontext.c b/lib/commands/toolcontext.c index c8d96ca..c150991 100644 --- a/lib/commands/toolcontext.c +++ b/lib/commands/toolcontext.c @@ -362,7 +362,8 @@ static void _init_logging(struct cmd_context *cmd) /* Tell device-mapper about our logging */ #ifdef DEVMAPPER_SUPPORT - dm_log_with_errno_init(print_log); + if (!dm_log_is_non_default()) + dm_log_with_errno_init(print_log); #endif reset_log_duplicated(); reset_lvm_errno(1);

1 0

master - thin: enhance logging
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=86b04ebd193c9db0… Commit: 86b04ebd193c9db09ee5f03a2de4997aa2ab52c3 Parent: 7e1c08bb6a90acdbf56e362ee017bd48de13a535 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Fri Oct 9 21:35:05 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 14:38:52 2015 +0200 thin: enhance logging Add debug message with more details about threshold overflow. --- lib/metadata/thin_manip.c | 16 ++++++++++++++-- 1 files changed, 14 insertions(+), 2 deletions(-) diff --git a/lib/metadata/thin_manip.c b/lib/metadata/thin_manip.c index 24c68a8..37bf4cb 100644 --- a/lib/metadata/thin_manip.c +++ b/lib/metadata/thin_manip.c @@ -224,15 +224,27 @@ int pool_below_threshold(const struct lv_segment *pool_seg) if (!lv_thin_pool_percent(pool_seg->lv, 0, &percent)) return_0; - if (percent >= threshold) + if (percent >= threshold) { + log_debug("Threshold configured for free data space in " + "thin pool %s has been reached (%.2f%% >= %.2f%%).", + display_lvname(pool_seg->lv), + dm_percent_to_float(percent), + dm_percent_to_float(threshold)); return 0; + } /* Metadata */ if (!lv_thin_pool_percent(pool_seg->lv, 1, &percent)) return_0; - if (percent >= threshold) + if (percent >= threshold) { + log_debug("Threshold configured for free metadata space in " + "thin pool %s has been reached (%.2f%% > %.2f%%).", + display_lvname(pool_seg->lv), + dm_percent_to_float(percent), + dm_percent_to_float(threshold)); return 0; + } return 1; }

1 0

master - dmeventd: correct ret code for fail
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=7e1c08bb6a90acdb… Commit: 7e1c08bb6a90acdbf56e362ee017bd48de13a535 Parent: c7b4359ff41634096770d464d362d43d2e82bbda Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 09:49:59 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 14:38:49 2015 +0200 dmeventd: correct ret code for fail As _do_register_device() calls register_device() function that returns 1 success & 0 fail - ret = -ENOMEM for fail case instead of 0. --- WHATS_NEW_DM | 1 + daemons/dmeventd/dmeventd.c | 6 ++++-- 2 files changed, 5 insertions(+), 2 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index e760875..a1c121a 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Return -ENOMEM when device registration fails instead of 0 (=success). Enforce writethrough mode for cleaner policy. Add support for recognition and deactivation of MD devices to blkdeactivate. Move target status functions out of libdm-deptree. diff --git a/daemons/dmeventd/dmeventd.c b/daemons/dmeventd/dmeventd.c index 1ff5bf9..ffc941f 100644 --- a/daemons/dmeventd/dmeventd.c +++ b/daemons/dmeventd/dmeventd.c @@ -1070,8 +1070,10 @@ static int _register_for_event(struct message_data *message_data) if (!(thread = _lookup_thread_status(message_data))) { _unlock_mutex(); - if (!(ret = _do_register_device(thread_new))) - goto out; + if (!_do_register_device(thread_new)) { + ret = -ENOMEM; + goto_out; + } thread = thread_new; thread_new = NULL;

1 0

master - thin: check for overflown pool earlier
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=c7b4359ff4163409… Commit: c7b4359ff41634096770d464d362d43d2e82bbda Parent: 5695c6aca63b2312e76f8b3fbbea95601269f8c8 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Wed Oct 7 15:10:03 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 14:37:07 2015 +0200 thin: check for overflown pool earlier Check for pool early before we actually start to modify metadata. This requires locally active thin-pool. --- WHATS_NEW | 1 + lib/metadata/lv_manip.c | 18 ++++++++++++------ 2 files changed, 13 insertions(+), 6 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index f683456..348c099 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Check for enough space in thin-pool in command before creating new thin. Make libblkid detect all copies of the same signature if use_blkid_wiping=1. Fix vgimportclone with -n to not add number unnecessarily to base VG name. Cleanup vgimportclone script and remove dependency on awk, grep, cut and tr. diff --git a/lib/metadata/lv_manip.c b/lib/metadata/lv_manip.c index 3e870d5..8a0bbcd 100644 --- a/lib/metadata/lv_manip.c +++ b/lib/metadata/lv_manip.c @@ -7087,12 +7087,18 @@ static struct logical_volume *_lv_create_an_lv(struct volume_group *vg, } } - if (seg_is_thin_volume(lp) && - lv_is_new_thin_pool(pool_lv)) { - thin_pool_was_active = lv_is_active(pool_lv); - if (!check_new_thin_pool(pool_lv)) - return_NULL; - /* New pool is now inactive */ + if (seg_is_thin_volume(lp)) { + if (lv_is_new_thin_pool(pool_lv)) { + thin_pool_was_active = lv_is_active(pool_lv); + if (!check_new_thin_pool(pool_lv)) + return_NULL; + /* New pool is now inactive */ + } else if (!pool_below_threshold(first_seg(pool_lv))) { + log_error("Cannot create new thin volume, free space in " + "thin pool %s reached threshold.", + display_lvname(pool_lv)); + return NULL; + } } if (seg_is_cache(lp) &&

1 0

master - libdm: enforce writethrough mode for cleaner
by Zdenek Kabelac 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=5695c6aca63b2312… Commit: 5695c6aca63b2312e76f8b3fbbea95601269f8c8 Parent: 5ac81657e5d312b72d746bc7cf22a4e6bb5b1609 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 13 14:06:15 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 13 14:35:48 2015 +0200 libdm: enforce writethrough mode for cleaner With "cleaner" policy always use 'writethrough' mode. --- WHATS_NEW_DM | 1 + libdm/libdm-deptree.c | 3 ++- 2 files changed, 3 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index e89ead0..e760875 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Enforce writethrough mode for cleaner policy. Add support for recognition and deactivation of MD devices to blkdeactivate. Move target status functions out of libdm-deptree. Correct use of max_write_behind parameter when generating raid target line. diff --git a/libdm/libdm-deptree.c b/libdm/libdm-deptree.c index ef283e9..afdce8d 100644 --- a/libdm/libdm-deptree.c +++ b/libdm/libdm-deptree.c @@ -3302,7 +3302,8 @@ int dm_tree_node_add_cache_target(struct dm_tree_node *node, return_0; seg->data_block_size = data_block_size; - seg->flags = feature_flags; + /* Enforce WriteThough mode for cleaner policy */ + seg->flags = (strcmp(policy_name, "cleaner") == 0) ? DM_CACHE_FEATURE_WRITETHROUGH : feature_flags; seg->policy_name = policy_name; /* FIXME: better validation missing */

1 0

master - wiping: make libblkid detect all copies of the same signature if use_blkid_wiping=1
by Peter Rajnoha 13 Oct '15

13 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=5ac81657e5d312b7… Commit: 5ac81657e5d312b72d746bc7cf22a4e6bb5b1609 Parent: 5bd63df237bc3426d1ebd41b103430063846b3ab Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Tue Oct 13 12:18:04 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Tue Oct 13 12:22:09 2015 +0200 wiping: make libblkid detect all copies of the same signature if use_blkid_wiping=1 Some signatures are spread around the disk in several copies, mainly for backup. Make libblkid to detect these extra copies - there was missing "blkid_probe_step_back" fn call after successful wipe of previous signature copy. An example with FAT table which has copies: $ mkfs.vfat /dev/sda1 Before this patch: $ pvcreate /dev/sda1 WARNING: vfat signature detected on /dev/sda1 at offset 54. Wipe it? [y/n]: y Wiping vfat signature on /dev/sda1. Physical volume "/dev/sda1" successfully created With this patch applied: $ pvcreate /dev/sda1 WARNING: vfat signature detected on /dev/sda1 at offset 54. Wipe it? [y/n]: y Wiping vfat signature on /dev/sda1. WARNING: vfat signature detected on /dev/sda1 at offset 0. Wipe it? [y/n]: y Wiping vfat signature on /dev/sda1. WARNING: vfat signature detected on /dev/sda1 at offset 510. Wipe it? [y/n]: y Wiping vfat signature on /dev/sda1. Physical volume "/dev/sda1" successfully created --- WHATS_NEW | 1 + lib/device/dev-type.c | 7 ++++++- 2 files changed, 7 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 5fa6e7f..f683456 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Make libblkid detect all copies of the same signature if use_blkid_wiping=1. Fix vgimportclone with -n to not add number unnecessarily to base VG name. Cleanup vgimportclone script and remove dependency on awk, grep, cut and tr. Add vg_missing_pv_count report field to report number of missing PVs in a VG. diff --git a/lib/device/dev-type.c b/lib/device/dev-type.c index 25f11d1..1d03024 100644 --- a/lib/device/dev-type.c +++ b/lib/device/dev-type.c @@ -651,8 +651,13 @@ static int _wipe_known_signatures_with_blkid(struct device *dev, const char *nam BLKID_SUBLKS_BADCSUM); while (!blkid_do_probe(probe)) { - if ((r_wipe = _blkid_wipe(probe, dev, name, types_to_exclude, types_no_prompt, yes, force)) == 1) + if ((r_wipe = _blkid_wipe(probe, dev, name, types_to_exclude, types_no_prompt, yes, force)) == 1) { (*wiped)++; + if (blkid_probe_step_back(probe)) { + log_error("Failed to step back blkid probe to check just wiped signature."); + goto out; + } + } /* do not count excluded types */ if (r_wipe != 2) found++;

1 0

master - tests: test vgimportclone and -n option
by Peter Rajnoha 12 Oct '15

12 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=5bd63df237bc3426… Commit: 5bd63df237bc3426d1ebd41b103430063846b3ab Parent: 75420282e17e44d5c7b56126a5c7fd170f0ef417 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Mon Oct 12 12:54:35 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Mon Oct 12 12:55:39 2015 +0200 tests: test vgimportclone and -n option --- test/shell/vgimportclone.sh | 25 +++++++++++++++++++++++++ 1 files changed, 25 insertions(+), 0 deletions(-) diff --git a/test/shell/vgimportclone.sh b/test/shell/vgimportclone.sh index 0b1af38..53c8b00 100644 --- a/test/shell/vgimportclone.sh +++ b/test/shell/vgimportclone.sh @@ -71,3 +71,28 @@ lvchange -ay $vg1/$lv1 $vg2/$lv1 vgchange -an $vg1 $vg2 vgremove -ff $vg1 $vg2 + +# Verify that if we provide the -n|--basevgname, +# the number suffix is not added unnecessarily. +vgcreate --metadatasize 128k A${vg1}B "$dev1" + +# vg1B is not the same as Avg1B - we don't need number suffix +dd if="$dev1" of="$dev2" bs=256K count=1 +aux notify_lvmetad "$dev2" +vgimportclone -n ${vg1}B "$dev2" +check pv_field "$dev2" vg_name ${vg1}B + +# Avg1 is not the same as Avg1B - we don't need number suffix +dd if="$dev1" of="$dev2" bs=256K count=1 +aux notify_lvmetad "$dev2" +vgimportclone -n A${vg1} "$dev2" +check pv_field "$dev2" vg_name A${vg1} + +# Avg1B is the same as Avg1B - we need to add the number suffix +dd if="$dev1" of="$dev2" bs=256K count=1 +aux notify_lvmetad "$dev2" +vgimportclone -n A${vg1}B "$dev2" +aux vgs +check pv_field "$dev2" vg_name A${vg1}B1 + +vgremove -ff A${vg1}B A${vg1}B1

1 0

master - vgimportclone: make sure there's no prefix in VG name list
by Peter Rajnoha 12 Oct '15

12 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=75420282e17e44d5… Commit: 75420282e17e44d5c7b56126a5c7fd170f0ef417 Parent: 38df48d108e3b040501edfd9e0517bfcb31373de Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Mon Oct 12 12:51:13 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Mon Oct 12 12:54:59 2015 +0200 vgimportclone: make sure there's no prefix in VG name list Make sure log/prefix is set to "" when getting the list of VG names. We need this for the format to be correct so it's properly searched through later on. --- scripts/vgimportclone.sh | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/scripts/vgimportclone.sh b/scripts/vgimportclone.sh index 52f3884..0ac987a 100755 --- a/scripts/vgimportclone.sh +++ b/scripts/vgimportclone.sh @@ -220,7 +220,7 @@ fi ### :vgname1:vgname2:...:vgnameN: ##################################################################### -OLDVGS=":`"${LVM}" vgs ${LVM_OPTS} -o name --noheadings --rows --separator :`:" +OLDVGS=":`"${LVM}" vgs ${LVM_OPTS} -o name --noheadings --rows --separator : --config 'log{prefix=""}'`:" checkvalue $? "Current VG names could not be collected without errors" #####################################################################

1 0

master - vgimportclone: fix vgimportclone with -n to not add number unnecessarily to base VG name
by Peter Rajnoha 12 Oct '15

12 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=38df48d108e3b040… Commit: 38df48d108e3b040501edfd9e0517bfcb31373de Parent: 21a8ac0cd3a392feaa049ab509c4727eee548d6b Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Mon Oct 12 11:11:34 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Mon Oct 12 11:11:34 2015 +0200 vgimportclone: fix vgimportclone with -n to not add number unnecessarily to base VG name $ vgcreate vgA /dev/sda Volume group "vgA" successfully created $ dd if=/dev/sda of=/dev/sdb bs=1M $ dd if=/dev/sda of=/dev/sdc bs=1M (the new VG name is prefix of existing VG name) $ vgimportclone -n vg /dev/sdb (the new VG name is suffix of existing VG name) $ vgimportclone -n gA /dev/sdc Before this patch: ------------------ (we end up with "vg1" and "gA1" names with the "1" suffix which is not needed) $ vgs -o vg_name VG gA1 vg1 vgA With this patch applied: ------------------------ (we end up with "vg" and "gA" names as they're unique already and no extra suffix is added) $ # vgs -o vg_name VG gA vg vgA Of course, if the name supplied is not unique, the number is added correctly: $ dd if=/dev/sda of=/dev/sdb bs=1M $ vgimportclone -n vgA /dev/sdb $ vgs -o vg_name VG vgA vgA1 --- WHATS_NEW | 1 + scripts/vgimportclone.sh | 8 +++++--- 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 9af759e..5fa6e7f 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Fix vgimportclone with -n to not add number unnecessarily to base VG name. Cleanup vgimportclone script and remove dependency on awk, grep, cut and tr. Add vg_missing_pv_count report field to report number of missing PVs in a VG. Properly identify internal LV holding sanlock locks within lv_role field. diff --git a/scripts/vgimportclone.sh b/scripts/vgimportclone.sh index d14a170..52f3884 100755 --- a/scripts/vgimportclone.sh +++ b/scripts/vgimportclone.sh @@ -48,7 +48,7 @@ function getvgname { NAME="${BNAME}" I=0 - while [[ "${VGLIST}" =~ "${NAME}" ]] + while [[ "${VGLIST}" =~ ":${NAME}:" ]] do I=$(($I+1)) NAME="${BNAME}$I" @@ -215,10 +215,12 @@ then fi ##################################################################### -### Get the existing state so we can use it later +### Get the existing state so we can use it later. +### The list of VG names is saved in this format: +### :vgname1:vgname2:...:vgnameN: ##################################################################### -OLDVGS=`"${LVM}" vgs ${LVM_OPTS} -o name --noheadings` +OLDVGS=":`"${LVM}" vgs ${LVM_OPTS} -o name --noheadings --rows --separator :`:" checkvalue $? "Current VG names could not be collected without errors" #####################################################################

1 0

master - Fix segfault when lvmlockd is running but not lvmetad
by David Teigland 09 Oct '15

09 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=21a8ac0cd3a392fe… Commit: 21a8ac0cd3a392feaa049ab509c4727eee548d6b Parent: 1f30ba6178df6c3faac06328f17158f549b9d4f7 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Fri Oct 9 12:20:22 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Fri Oct 9 12:20:22 2015 -0500 Fix segfault when lvmlockd is running but not lvmetad If lvmlockd is running, lvmetad is configured (use_lvmetad=1), but lvmetad is not running, then commands will seg fault when trying to send a message to lvmetad. The difference is lvmetad being "active", not just "used". --- lib/cache/lvmetad.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/lib/cache/lvmetad.c b/lib/cache/lvmetad.c index 20d2001..c7f6ecb 100644 --- a/lib/cache/lvmetad.c +++ b/lib/cache/lvmetad.c @@ -1683,7 +1683,7 @@ void lvmetad_validate_global_cache(struct cmd_context *cmd, int force) return; } - if (!lvmetad_used()) + if (!lvmetad_active()) return; log_debug_lvmetad("Validating global lvmetad cache");

1 0

master - vgimportclone: cleanup the script and remove dependency on awk, grep, cut and tr
by Peter Rajnoha 09 Oct '15

09 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=1f30ba6178df6c3f… Commit: 1f30ba6178df6c3faac06328f17158f549b9d4f7 Parent: 8733a8d89040820f66a99d4d4e8f524ab2497983 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 9 16:26:27 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 9 16:31:02 2015 +0200 vgimportclone: cleanup the script and remove dependency on awk, grep, cut and tr We can replace the expressions with awk/grep/cut/tr with --select now and more suitable reporting options and modes. Also, we don't need to check the temporary lvm.conf generated within vgimportclone script since we're generating it ourselves now using lvmconfig, not using sed anymore like it was before (so we can be pretty sure it's correct - we use lvmconfig now even for generating the lvm.conf itself). --- WHATS_NEW | 1 + scripts/vgimportclone.sh | 87 +++++++++++++-------------------------------- 2 files changed, 26 insertions(+), 62 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index ac9f6b8..9af759e 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Cleanup vgimportclone script and remove dependency on awk, grep, cut and tr. Add vg_missing_pv_count report field to report number of missing PVs in a VG. Properly identify internal LV holding sanlock locks within lv_role field. Add metadata_devices and seg_metadata_le_ranges report fields for raid vols. diff --git a/scripts/vgimportclone.sh b/scripts/vgimportclone.sh index af8cf5c..d14a170 100755 --- a/scripts/vgimportclone.sh +++ b/scripts/vgimportclone.sh @@ -21,11 +21,7 @@ RM=rm BASENAME=basename MKTEMP=mktemp -AWK=awk -CUT=cut -TR=tr READLINK=readlink -GREP=grep GETOPT=getopt # user may override lvm location by setting LVM_BINARY @@ -253,59 +249,37 @@ checkvalue $? "Failed to generate ${LVMCONF}" # Only keep TMP_LVM_SYSTEM_DIR if it contains something worth keeping [ -n "${DEBUG}" ] && KEEP_TMP_LVM_SYSTEM_DIR=1 -# verify the config contains the filter, scan and cache_dir (or cache) config keywords -"$GREP" -q '^[[:space:]]*filter[[:space:]]*=' ${LVMCONF} || \ - die 5 "Temporary lvm.conf must contain 'filter' config." -"$GREP" -q '^[[:space:]]*scan[[:space:]]*=' ${LVMCONF} || \ - die 6 "Temporary lvm.conf must contain 'scan' config." - -# check for either 'cache' or 'cache_dir' config values -"$GREP" -q '[[:space:]]*cache[[:space:]]*=' ${LVMCONF} -CACHE_RET=$? -"$GREP" -q '^[[:space:]]*cache_dir' ${LVMCONF} -CACHE_DIR_RET=$? -[ $CACHE_RET -eq 0 -o $CACHE_DIR_RET -eq 0 ] || \ - die 7 "Temporary lvm.conf must contain 'cache' or 'cache_dir' config." - ### set to use new lvm.conf export LVM_SYSTEM_DIR=${TMP_LVM_SYSTEM_DIR} +# Check if there are any PVs that don't belong to any VG +# or even if there are disks which are not PVs at all. +NOVGDEVLIST=`${LVM} pvs -a -o pv_name --select vg_name="" --noheadings` +checkvalue $? "Failed to collect information for PV check" +if [ -n "${NOVGDEVLIST}" ]; then + FOLLOWLIST="" + while read PVNAME; do + FOLLOW=`$READLINK $PVNAME` + FOLLOWLIST="$FOLLOWLIST $FOLLOW" + done <<< "`echo "${NOVGDEVLIST}"`" + die 8 "Specified devices don't belong to a VG:$FOLLOWLIST" +fi ##################################################################### ### Rename the VG(s) and change the VG and PV UUIDs. ##################################################################### +VGLIST=`${LVM} vgs -o vg_name,vg_exported,vg_missing_pv_count --noheadings --binary` +checkvalue $? "Failed to collect VG information" -PVINFO=`"${LVM}" pvs ${LVM_OPTS} -o pv_name,vg_name,vg_attr --noheadings --separator :` -checkvalue $? "PV info could not be collected without errors" - -# output VG info so each line looks like: name:exported?:disk1,disk2,... -VGINFO=`echo "${PVINFO}" | \ - "$AWK" -F : '{{sub(/^[ \t]*/,"")} \ - {sub(/unknown device/,"unknown_device")} \ - {vg[$2]=$1","vg[$2]} if($3 ~ /^..x/){x[$2]="x"}} \ - END{for(k in vg){printf("%s:%s:%s\n", k, x[k], vg[k])}}'` -checkvalue $? "PV info could not be parsed without errors" - -for VG in ${VGINFO} -do - VGNAME=`echo "${VG}" | "$CUT" -d: -f1` - EXPORTED=`echo "${VG}" | "$CUT" -d: -f2` - PVLIST=`echo "${VG}" | "$CUT" -d: -f3- | "$TR" , ' '` - - if [ -z "${VGNAME}" ] - then - FOLLOWLIST="" - for DEV in $PVLIST; do - FOLLOW=`"$READLINK" $DEV` - FOLLOWLIST="$FOLLOW $FOLLOWLIST" - done - die 8 "Specified PV(s) ($FOLLOWLIST) don't belong to a VG." +while read VGNAME VGEXPORTED VGMISSINGPVCOUNT; do + if [ $VGMISSINGPVCOUNT -gt 0 ]; then + echo "Volume Group ${VGNAME} has unknown PV(s), skipping." + echo "- Were all associated PV(s) supplied as arguments?" + continue fi - if [ -n "${EXPORTED}" ] - then - if [ ${IMPORT} -eq 1 ] - then + if [ "$VGEXPORTED" = "1" ]; then + if [ ${IMPORT} -eq 1 ]; then "$LVM" vgimport ${LVM_OPTS} ${TEST_OPT} "${VGNAME}" checkvalue $? "Volume Group ${VGNAME} could not be imported" else @@ -314,23 +288,12 @@ do fi fi - ### change the pv uuids - if [[ "${PVLIST}" =~ "unknown" ]] - then - echo "Volume Group ${VGNAME} has unknown PV(s), skipping." - echo "- Were all associated PV(s) supplied as arguments?" - continue - fi - - for BLOCKDEV in ${PVLIST} - do - "$LVM" pvchange ${LVM_OPTS} ${TEST_OPT} --uuid ${BLOCKDEV} --config 'global{activation=0}' - checkvalue $? "Unable to change PV uuid for ${BLOCKDEV}" - done + "$LVM" pvchange ${LVM_OPTS} ${TEST_OPT} --uuid --config 'global{activation=0}' --select "vg_name=${VGNAME}" + checkvalue $? "Unable to change all PV uuids in VG ${VG_NAME}" NEWVGNAME=`getvgname "${OLDVGS}" "${VGNAME}" "${NEWVG}"` - "$LVM" vgchange ${LVM_OPTS} ${TEST_OPT} --uuid "${VGNAME}" --config 'global{activation=0}' + "$LVM" vgchange ${LVM_OPTS} ${TEST_OPT} --uuid --config 'global{activation=0}' ${VGNAME} checkvalue $? "Unable to change VG uuid for ${VGNAME}" ## if the name isn't going to get changed dont even try. @@ -341,7 +304,7 @@ do fi CHANGES_MADE=1 -done +done <<< "`echo "${VGLIST}"`" ##################################################################### ### Restore the old environment

1 0

master - report: add vg_missing_pv_count field
by Peter Rajnoha 09 Oct '15

09 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=8733a8d89040820f… Commit: 8733a8d89040820f66a99d4d4e8f524ab2497983 Parent: 5446d177569ca10476c935013ae03e6a0763ab34 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Fri Oct 9 16:20:29 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Fri Oct 9 16:28:58 2015 +0200 report: add vg_missing_pv_count field We already have pv_count to report number of PVs that a VG has based on metadata. This patch exposes the information about how many of these PVs are missing which is also useful information for a VG. Wwe could count the sum of pv_missing reporting fields for each PV in the VG before, but the new field is practical when reporting VG as a whole and there's no need to process each PV from VG alone. --- WHATS_NEW | 1 + lib/report/columns.h | 3 ++- lib/report/properties.c | 2 ++ lib/report/report.c | 11 +++++++++++ 4 files changed, 16 insertions(+), 1 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 227530d..ac9f6b8 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Add vg_missing_pv_count report field to report number of missing PVs in a VG. Properly identify internal LV holding sanlock locks within lv_role field. Add metadata_devices and seg_metadata_le_ranges report fields for raid vols. Fix lvm2-{activation,clvmd,cmirrord,monitor} service to exec before mounting. diff --git a/lib/report/columns.h b/lib/report/columns.h index 6041eeb..41a03f4 100644 --- a/lib/report/columns.h +++ b/lib/report/columns.h @@ -158,7 +158,8 @@ FIELD(VGS, vg, NUM, "#Ext", extent_count, 4, uint32, vg_extent_count, "Total num FIELD(VGS, vg, NUM, "Free", free_count, 4, uint32, vg_free_count, "Total number of unallocated Physical Extents.", 0) FIELD(VGS, vg, NUM, "MaxLV", max_lv, 5, uint32, max_lv, "Maximum number of LVs allowed in VG or 0 if unlimited.", 0) FIELD(VGS, vg, NUM, "MaxPV", max_pv, 5, uint32, max_pv, "Maximum number of PVs allowed in VG or 0 if unlimited.", 0) -FIELD(VGS, vg, NUM, "#PV", pv_count, 3, uint32, pv_count, "Number of PVs.", 0) +FIELD(VGS, vg, NUM, "#PV", pv_count, 3, uint32, pv_count, "Number of PVs in VG.", 0) +FIELD(VGS, vg, NUM, "#PV Missing", cmd, 11, vgmissingpvcount, vg_missing_pv_count, "Number of PVs in VG which are missing.", 0) FIELD(VGS, vg, NUM, "#LV", cmd, 3, lvcount, lv_count, "Number of LVs.", 0) FIELD(VGS, vg, NUM, "#SN", cmd, 3, snapcount, snap_count, "Number of snapshots.", 0) FIELD(VGS, vg, NUM, "Seq", seqno, 3, uint32, vg_seqno, "Revision number of internal metadata. Incremented whenever it changes.", 0) diff --git a/lib/report/properties.c b/lib/report/properties.c index 262cc26..7dafbe5 100644 --- a/lib/report/properties.c +++ b/lib/report/properties.c @@ -420,6 +420,8 @@ GET_VG_NUM_PROPERTY_FN(vg_mda_copies, (vg_mda_copies(vg))) SET_VG_NUM_PROPERTY_FN(vg_mda_copies, vg_set_mda_copies) GET_VG_STR_PROPERTY_FN(vg_profile, vg_profile_dup(vg)) #define _vg_profile_set prop_not_implemented_set +GET_VG_NUM_PROPERTY_FN(vg_missing_pv_count, vg_missing_pv_count(vg)) +#define _vg_missing_pv_count_set prop_not_implemented_set /* LVSEG */ GET_LVSEG_STR_PROPERTY_FN(segtype, lvseg_segtype_dup(lvseg->lv->vg->vgmem, lvseg)) diff --git a/lib/report/report.c b/lib/report/report.c index 7ac1ac5..dea66ee 100644 --- a/lib/report/report.c +++ b/lib/report/report.c @@ -2438,6 +2438,17 @@ static int _vgprofile_disp(struct dm_report *rh, struct dm_pool *mem, return _field_set_value(field, "", NULL); } +static int _vgmissingpvcount_disp(struct dm_report *rh, struct dm_pool *mem, + struct dm_report_field *field, + const void *data, void *private) +{ + const struct volume_group *vg = (const struct volume_group *) data; + uint32_t count = vg_missing_pv_count(vg); + + return _uint32_disp(rh, mem, field, &count, private); +} + + static int _pvmdafree_disp(struct dm_report *rh, struct dm_pool *mem, struct dm_report_field *field, const void *data, void *private)

1 0

master - man lvmsystemid: fix typos
by David Teigland 08 Oct '15

08 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=5446d177569ca104… Commit: 5446d177569ca10476c935013ae03e6a0763ab34 Parent: c9ff5c8223c6c2d23829aef076d58c32382891c8 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Thu Oct 8 13:21:42 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Thu Oct 8 13:26:09 2015 -0500 man lvmsystemid: fix typos --- man/lvmsystemid.7.in | 12 ++++++------ 1 files changed, 6 insertions(+), 6 deletions(-) diff --git a/man/lvmsystemid.7.in b/man/lvmsystemid.7.in index eb10ac5..0048873 100644 --- a/man/lvmsystemid.7.in +++ b/man/lvmsystemid.7.in @@ -274,7 +274,7 @@ cache with pvscan --cache will allow a host to recognize the newly exported VG. vgimport sets the VG system_id to the local system_id as determined by -lvm.conf system_id_sources. vgimport automatically scans storage for +lvm.conf system_id_source. vgimport automatically scans storage for newly exported VGs. After vgimport, the exporting host will continue to see the VG as @@ -300,9 +300,9 @@ To move a VG from one host to another, vgexport and vgimport should be used. To forcibly gain ownership of a foreign VG, a host can add the foreign -system_id to its allow_system_id list, change the system_id of the foreign -VG to its own, and remove the foreign system_id from its allow_system_id -list. +system_id to its extra_system_ids list, change the system_id of the +foreign VG to its own, and remove the foreign system_id from its +extra_system_ids list. .SS shared VGs @@ -328,7 +328,7 @@ creation_host will be the same. Orphan PVs are unused devices; they are not currently used in any VG. Because of this, they are not protected by a system_id, and any host can -use them. Coodination of changes to orphan PVs is beyond the scope of +use them. Coordination of changes to orphan PVs is beyond the scope of system_id. The same is true of any block device that is not a PV. The effects of this are especially evident when lvm uses lvmetad caching. @@ -337,7 +337,7 @@ using the orphan, the other hosts will continue to report the PV as an orphan. Nothing would automatically prevent the other hosts from using the newly allocated PV and corrupting it. If the other hosts run a command to rescan devices, and update lvmetad, they would then recognize -the PV has been used by another host. A command that rescans devices +that the PV has been used by another host. A command that rescans devices could be pvscan --cache, or vgs --foreign. .SH SEE ALSO

1 0

master - man lvmsystemd: fix typo
by David Teigland 08 Oct '15

08 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=c9ff5c8223c6c2d2… Commit: c9ff5c8223c6c2d23829aef076d58c32382891c8 Parent: d99dd4086da5ab2ed7bf02a8711d4a4998fe2b31 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Thu Oct 8 13:14:28 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Thu Oct 8 13:14:28 2015 -0500 man lvmsystemd: fix typo --- man/lvmsystemid.7.in | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/man/lvmsystemid.7.in b/man/lvmsystemid.7.in index 37a01af..eb10ac5 100644 --- a/man/lvmsystemid.7.in +++ b/man/lvmsystemid.7.in @@ -84,7 +84,7 @@ version without the system_id feature. .SS Types of VG access -A local VG is mean to be used by a single host. +A local VG is meant to be used by a single host. .br A shared or clustered VG is meant to be used by multiple hosts. .br

1 0

master - vgcreate: improve error message for multiple lock managers
by David Teigland 08 Oct '15

08 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=d99dd4086da5ab2e… Commit: d99dd4086da5ab2ed7bf02a8711d4a4998fe2b31 Parent: 09981afc1c03e1d79109c483b3d19ae1c207e158 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Thu Oct 8 10:38:35 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Thu Oct 8 10:41:13 2015 -0500 vgcreate: improve error message for multiple lock managers If 'vgcreate --shared' finds both sanlock and dlm are running, print a more accurate error message: "Found multiple lock managers, select one with --lock-type." When neither is running, we still print: "Failed to detect a running lock manager to select lock type." --- lib/locking/lvmlockd.c | 5 ++--- lib/locking/lvmlockd.h | 4 ++-- tools/toollib.c | 9 +++++++-- 3 files changed, 11 insertions(+), 7 deletions(-) diff --git a/lib/locking/lvmlockd.c b/lib/locking/lvmlockd.c index f741940..2bc34c8 100644 --- a/lib/locking/lvmlockd.c +++ b/lib/locking/lvmlockd.c @@ -2561,7 +2561,7 @@ int lockd_rename_vg_final(struct cmd_context *cmd, struct volume_group *vg, int return 1; } -const char *lockd_running_lock_type(struct cmd_context *cmd) +const char *lockd_running_lock_type(struct cmd_context *cmd, int *found_multiple) { daemon_reply reply; const char *lock_type = NULL; @@ -2583,10 +2583,9 @@ const char *lockd_running_lock_type(struct cmd_context *cmd) switch (result) { case -EXFULL: - log_error("lvmlockd found multiple lock managers, use --lock-type to select one."); + *found_multiple = 1; break; case -ENOLCK: - log_error("lvmlockd found no lock manager running."); break; case LOCK_TYPE_SANLOCK: log_debug("lvmlockd found sanlock"); diff --git a/lib/locking/lvmlockd.h b/lib/locking/lvmlockd.h index 7e6505d..8b282d8 100644 --- a/lib/locking/lvmlockd.h +++ b/lib/locking/lvmlockd.h @@ -89,7 +89,7 @@ int lockd_init_lv_args(struct cmd_context *cmd, struct volume_group *vg, int lockd_free_lv(struct cmd_context *cmd, struct volume_group *vg, const char *lv_name, struct id *lv_id, const char *lock_args); -const char *lockd_running_lock_type(struct cmd_context *cmd); +const char *lockd_running_lock_type(struct cmd_context *cmd, int *found_multiple); int handle_sanlock_lv(struct cmd_context *cmd, struct volume_group *vg); @@ -223,7 +223,7 @@ static inline int lockd_free_lv(struct cmd_context *cmd, struct volume_group *vg return 1; } -static inline const char *lockd_running_lock_type(struct cmd_context *cmd) +static inline const char *lockd_running_lock_type(struct cmd_context *cmd, int *found_multiple) { log_error("Using a shared lock type requires lvmlockd."); return NULL; diff --git a/tools/toollib.c b/tools/toollib.c index cf5e236..dfb2b87 100644 --- a/tools/toollib.c +++ b/tools/toollib.c @@ -931,9 +931,14 @@ int vgcreate_params_set_from_args(struct cmd_context *cmd, } } else if (arg_is_set(cmd, shared_ARG)) { + int found_multiple = 0; + if (use_lvmlockd) { - if (!(lock_type = lockd_running_lock_type(cmd))) { - log_error("Failed to detect a running lock manager to select lock type."); + if (!(lock_type = lockd_running_lock_type(cmd, &found_multiple))) { + if (found_multiple) + log_error("Found multiple lock managers, select one with --lock-type."); + else + log_error("Failed to detect a running lock manager to select lock type."); return 0; }

1 0

master - vgcreate: allow both --shared and --lock-type options
by David Teigland 08 Oct '15

08 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=09981afc1c03e1d7… Commit: 09981afc1c03e1d79109c483b3d19ae1c207e158 Parent: 3d03e504cda6307f692c5879f6850b2e27af7d71 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Thu Oct 8 10:14:33 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Thu Oct 8 10:14:33 2015 -0500 vgcreate: allow both --shared and --lock-type options Using --lock-type sanlock|dlm implies --shared. Using --shared selects lock type sanlock|dlm (by choosing the one that's running.) Using both --shared and --lock-type sanlock|dlm should also be allowed (--shared is just redundant information.) --- tools/toollib.c | 10 ++++++++-- 1 files changed, 8 insertions(+), 2 deletions(-) diff --git a/tools/toollib.c b/tools/toollib.c index 01146a4..cf5e236 100644 --- a/tools/toollib.c +++ b/tools/toollib.c @@ -898,12 +898,18 @@ int vgcreate_params_set_from_args(struct cmd_context *cmd, use_clvmd = (locking_type == 3); if (arg_is_set(cmd, locktype_ARG)) { - if (arg_is_set(cmd, clustered_ARG) || arg_is_set(cmd, shared_ARG)) { - log_error("A lock type cannot be specified with --shared or --clustered."); + if (arg_is_set(cmd, clustered_ARG)) { + log_error("A lock type cannot be specified with --clustered."); return 0; } + lock_type = arg_str_value(cmd, locktype_ARG, ""); + if (arg_is_set(cmd, shared_ARG) && !is_lockd_type(lock_type)) { + log_error("The --shared option requires lock type sanlock or dlm."); + return 0; + } + } else if (arg_is_set(cmd, clustered_ARG)) { const char *arg_str = arg_str_value(cmd, clustered_ARG, ""); int clustery = strcmp(arg_str, "y") ? 0 : 1;

1 0

master - metadata: format_text: provide more detailed error message when metadata too large for PV mda
by Peter Rajnoha 08 Oct '15

08 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=3d03e504cda6307f… Commit: 3d03e504cda6307f692c5879f6850b2e27af7d71 Parent: e04424e87e66df22578d1e4d2488615cd3692873 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Thu Oct 8 16:25:10 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Thu Oct 8 16:27:03 2015 +0200 metadata: format_text: provide more detailed error message when metadata too large for PV mda Also, leave out the note about "circular buffer" which is an internal imeplementation detail anyway and not quite informational for users: Before this patch: $ vgcreate vg1 /dev/sda VG vg1 metadata too large for circular buffer Failed to write VG vg1. With this patch applied: $ vgcreate vg1 /dev/sda VG vg1 metadata too large: size of metadata to write is 691 bytes while PV metadata area size on /dev/sda is 512 bytes. Failed to write VG vg1. --- lib/format_text/format-text.c | 7 +++++-- 1 files changed, 5 insertions(+), 2 deletions(-) diff --git a/lib/format_text/format-text.c b/lib/format_text/format-text.c index e0ec8ed..c1c4787 100644 --- a/lib/format_text/format-text.c +++ b/lib/format_text/format-text.c @@ -655,8 +655,11 @@ static int _vg_write_raw(struct format_instance *fid, struct volume_group *vg, if ((new_wrap && old_wrap) || (rlocn && (new_wrap || old_wrap) && (new_end > rlocn->offset)) || (mdac->rlocn.size >= mdah->size)) { - log_error("VG %s metadata too large for circular buffer", - vg->name); + log_error("VG %s metadata too large: size of metadata to write " + "is %" PRIu64 " bytes while PV metadata area size " + "on %s is %" PRIu64 " bytes.", + vg->name, mdac->rlocn.size, + dev_name(mdac->area.dev), mdah->size); goto out; }

1 0

master - report: identify LV hodling sanlock locks as 'private, lockd, sanlock' within lv_role report field
by Peter Rajnoha 08 Oct '15

08 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=e04424e87e66df22… Commit: e04424e87e66df22578d1e4d2488615cd3692873 Parent: 277dd0aa7a9013697ae99164900d8d8ec12a2850 Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Thu Oct 8 13:44:21 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Thu Oct 8 13:44:29 2015 +0200 report: identify LV hodling sanlock locks as 'private,lockd,sanlock' within lv_role report field Before this patch: $ lvs -a -o name,layout,role test/lvmlock LV Layout Role [lvmlock] linear public With this patch applied: $ lvs -a -o name,layout,role test/lvmlock LV Layout Role [lvmlock] linear private,lockd,sanlock --- WHATS_NEW | 1 + lib/metadata/lv_manip.c | 11 +++++++++++ 2 files changed, 12 insertions(+), 0 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index 5da1552..227530d 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Properly identify internal LV holding sanlock locks within lv_role field. Add metadata_devices and seg_metadata_le_ranges report fields for raid vols. Fix lvm2-{activation,clvmd,cmirrord,monitor} service to exec before mounting. diff --git a/lib/metadata/lv_manip.c b/lib/metadata/lv_manip.c index 48d7d34..3e870d5 100644 --- a/lib/metadata/lv_manip.c +++ b/lib/metadata/lv_manip.c @@ -129,6 +129,8 @@ enum { LV_TYPE_RAID6_ZR, LV_TYPE_RAID6_NR, LV_TYPE_RAID6_NC, + LV_TYPE_LOCKD, + LV_TYPE_SANLOCK }; static const char *_lv_type_names[] = { @@ -173,6 +175,8 @@ static const char *_lv_type_names[] = { [LV_TYPE_RAID6_ZR] = SEG_TYPE_NAME_RAID6_ZR, [LV_TYPE_RAID6_NR] = SEG_TYPE_NAME_RAID6_NR, [LV_TYPE_RAID6_NC] = SEG_TYPE_NAME_RAID6_NC, + [LV_TYPE_LOCKD] = "lockd", + [LV_TYPE_SANLOCK] = "sanlock", }; static int _lv_layout_and_role_mirror(struct dm_pool *mem, @@ -505,6 +509,13 @@ int lv_layout_and_role(struct dm_pool *mem, const struct logical_volume *lv, if (!_lv_layout_and_role_thick_origin_snapshot(mem, lv, *layout, *role, &public_lv)) goto_bad; + if (lv_is_lockd_sanlock_lv(lv)) { + if (!str_list_add_no_dup_check(mem, *role, _lv_type_names[LV_TYPE_LOCKD]) || + !str_list_add_no_dup_check(mem, *role, _lv_type_names[LV_TYPE_SANLOCK])) + goto_bad; + public_lv = 0; + } + /* * If layout not yet determined, it must be either * linear or striped or mixture of these two.

1 0

master - tests: check devices with prefix aren't in use
by Zdenek Kabelac 06 Oct '15

06 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=277dd0aa7a901369… Commit: 277dd0aa7a9013697ae99164900d8d8ec12a2850 Parent: ded9452174664b7e204f5ce266bbd5c770714474 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 6 14:55:24 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 6 15:23:55 2015 +0200 tests: check devices with prefix aren't in use Avoid running tests, when prefix already exist in the system. As prefix just uses PID number, we may hit a case for long running tests, where devices from some previous runs were not properly cleared away - detect this and fail early. (Such machine should be inspected and fixed). --- test/lib/inittest.sh | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/test/lib/inittest.sh b/test/lib/inittest.sh index a5eeb57..4ed8d20 100644 --- a/test/lib/inittest.sh +++ b/test/lib/inittest.sh @@ -29,6 +29,9 @@ TESTOLDPWD=$(pwd) COMMON_PREFIX="LVMTEST" PREFIX="${COMMON_PREFIX}$$" +# Check we are not conflickting with some exiting setup +dmsetup table | not grep "$PREFIX" || die "DM table already has devices with prefix $PREFIX!" + if test -z "$LVM_TEST_DIR"; then LVM_TEST_DIR=$TMPDIR; fi TESTDIR=$(mkdtemp "${LVM_TEST_DIR:-/tmp}" "$PREFIX.XXXXXXXXXX") || \ die "failed to create temporary directory in ${LVM_TEST_DIR:-$TESTOLDPWD}"

1 0

master - man: lvcreate continue
by Zdenek Kabelac 06 Oct '15

06 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=ded9452174664b7e… Commit: ded9452174664b7e204f5ce266bbd5c770714474 Parent: 4b1cadbd87a0814a46b36b07cfe94eda7baacb28 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Tue Oct 6 14:55:09 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 6 15:23:55 2015 +0200 man: lvcreate continue Finish remaing bits of updating pages for better rendering with -Thtml, -Tps. --- man/lvchange.8.in | 3 + man/lvcreate.8.in | 277 ++++++++++++++++++++++++++++++---------------------- man/lvm.8.in | 1 + 3 files changed, 164 insertions(+), 117 deletions(-) diff --git a/man/lvchange.8.in b/man/lvchange.8.in index 62bdca1..8ae17a0 100644 --- a/man/lvchange.8.in +++ b/man/lvchange.8.in @@ -431,7 +431,10 @@ Suppress locking failure messages. Changes the permission on volume lvol1 in volume group vg00 to be read-only: .sp .B lvchange \-pr vg00/lvol1 +. .SH SEE ALSO +. +.nh .BR lvm (8), .BR lvmetad (8), .BR lvs (8), diff --git a/man/lvcreate.8.in b/man/lvcreate.8.in index 8a364c4..2ce39dd 100644 --- a/man/lvcreate.8.in +++ b/man/lvcreate.8.in @@ -1,7 +1,22 @@ .TH LVCREATE 8 "LVM TOOLS #VERSION#" "Sistina Software UK" \" -*- nroff -*- +. +.\" Use 1st. parameter with \% to fix 'man2html' rendeing on same line! +.de SIZE_G +. IR \\$1 \c +. RB [ b | B | s | S | k | K | m | M | g | G ] +.. +.de SIZE_E +. IR \\$1 \c +. RB [ b | B | s | S | k | K | m | M | \c +. BR g | G | t | T | p | P | e | E ] +.. +. .SH NAME +. lvcreate \- create a logical volume in an existing volume group +. .SH SYNOPSIS +. .ad l .B lvcreate .RB [ \-a | \-\-activate @@ -17,7 +32,7 @@ lvcreate \- create a logical volume in an existing volume group .RB { passthrough | writeback | writethrough }] .RB [ \-\-cachepolicy .IR policy ] -.RB [ \-\-cachepool +.RB \%[ \-\-cachepool .IR CachePoolLogicalVolume ] .RB [ \-\-cachesettings .IR key \fB= value ] @@ -25,7 +40,7 @@ lvcreate \- create a logical volume in an existing volume group .IR ChunkSize ] .RB [ \-\-commandprofile .IR ProfileName ] -.RB [ \-C | \-\-contiguous +.RB \%[ \-C | \-\-contiguous .RB { y | n }] .RB [ \-d | \-\-debug ] .RB [ \-\-discards @@ -78,7 +93,7 @@ lvcreate \- create a logical volume in an existing volume group .IR Rate ] .RB [ \-r | \-\-readahead .RB { \fIReadAheadSectors | auto | none }] -.RB [ \-k | \-\-setactivationskip +.RB \%[ \-k | \-\-setactivationskip .RB { y | n }] .RB [ \-s | \-\-snapshot ] .RB [ \-V | \-\-virtualsize @@ -94,9 +109,10 @@ lvcreate \- create a logical volume in an existing volume group .RB { y | n }] .RB [ \-Z | \-\-zero .RB { y | n }] -.RI [ VolumeGroup [ \fB/ { ExternalOrigin |\: Origin | -.IR Pool } LogicalVolumeName ] -.RI [ PhysicalVolumePath [ \fB: \fIPE \fR[ \fB\- PE ]]...] +.RI [ VolumeGroup \c +.RB [ / \c +.RI { ExternalOrigin | Origin | Pool } LogicalVolumeName +.RI [ PhysicalVolumePath [ \fB: \fIPE \fR[ \fB\- PE ]]...]] .LP .B lvcreate .RB [ \-l | \-\-extents @@ -107,7 +123,7 @@ lvcreate \- create a logical volume in an existing volume group .IR LogicalVolumeSize ] .RB [ \-c | \-\-chunksize .IR ChunkSize ] -.RB [ \-\-commandprofile +.RB \%[ \-\-commandprofile .IR Profile\%Name ] .RB [ \-\-noudevsync ] .RB [ \-\-ignoremonitoring ] @@ -118,11 +134,13 @@ lvcreate \- create a logical volume in an existing volume group .RB [ \-n | \-\-name .IR SnapshotLogicalVolume ] .BR \-s | \-\-snapshot | \-H | \-\-cache -.RI \%{[ VolumeGroup \fB/ ] OriginalLogicalVolume -.BR \-V | \-\-virtualsize -.IR VirtualSize } +.RI \%{[ VolumeGroup \fB/\fP] OriginalLogicalVolume +.RB \%[ \-V | \-\-virtualsize +.IR VirtualSize ]} .ad b +. .SH DESCRIPTION +. lvcreate creates a new logical volume in a volume group (see .BR vgcreate "(8), " vgchange (8)) by allocating logical extents from the free physical extent pool @@ -151,40 +169,41 @@ for common options. .br Controls the availability of the Logical Volumes for immediate use after the command finishes running. -By default, new Logical Volumes are activated (\fB\-a\fIy\fR). -If it is possible technically, \fB\-a\fIn\fR will leave the new Logical +By default, new Logical Volumes are activated (\fB\-ay\fP). +If it is possible technically, \fB\-an\fP will leave the new Logical Volume inactive. But for example, snapshots of active origin can only be -created in the active state so \fB\-a\fIn\fR cannot be used with -\fB-\-type\fP \fIsnapshot\fP. This does not apply to thin volume snapshots, +created in the active state so \fB\-an\fP cannot be used with +\fB-\-type snapshot\fP. This does not apply to thin volume snapshots, which are by default created with flag to skip their activation -(\fB-k\fP\fIy\fP). -Normally the \fB\-\-zero\fP \fIn\fP argument has to be supplied too because +(\fB-ky\fP). +Normally the \fB\-\-zero n\fP argument has to be supplied too because zeroing (the default behaviour) also requires activation. -If autoactivation option is used (\fB\-a\fIay\fR), the logical volume is +If autoactivation option is used (\fB\-aay\fP), the logical volume is activated only if it matches an item in the -.IR activation / auto_activation_volume_list +\fBactivation/auto_activation_volume_list\fP set in \fBlvm.conf\fP(5). -For autoactivated logical volumes, \fB\-\-zero\fP \fIn\fP and -\fB\-\-wipesignatures\fP \fIn\fP is always assumed and it can't +For autoactivated logical volumes, \fB\-\-zero n\fP and +\fB\-\-wipesignatures n\fP is always assumed and it can't be overridden. If the clustered locking is enabled, -\fB\-a\fIey\fR will activate exclusively on one node and -.IR \fB\-a { a | l } y +\fB\-aey\fP will activate exclusively on one node and +.BR \-a { a | l } y will activate only on the local node. . .HP .BR \-H | \-\-cache .br -Creates cache or cache pool logical volume or both. -Specifying the optional argument \fB\-\-size\fP will cause the creation of -the cache logical volume. +Creates cache or cache pool logical volume. +.\" or both. +Specifying the optional argument \fB\-\-extents\fP or \fB\-\-size\fP +will cause the creation of the cache logical volume. .\" Specifying the optional argument \fB\-\-pooldatasize\fP will cause .\" the creation of the cache pool logical volume. -Specifying both arguments will cause the creation of cache with its -cache pool volume. +.\" Specifying both arguments will cause the creation of cache with its +.\" cache pool volume. When the Volume group name is specified together with existing logical volume -name which is NOT a cache pool name, such volume is treaded -as cache origin volume and cache pool is created. In this case -the \fB\-\-size\fP is used to specify size of cache pool volume. +name which is NOT a cache pool name, such volume is treated +as cache origin volume and cache pool is created. In this case the +\fB\-\-extents\fP or \fB\-\-size\fP is used to specify size of cache pool volume. See \fBlvmcache\fP(7) for more info about caching support. Note that the cache segment type requires a dm-cache kernel module version 1.3.0 or greater. @@ -228,19 +247,18 @@ and removes it from the list of settings stored in lvm2 metadata. . .HP .BR \-c | \-\-chunksize -.IR ChunkSize \c -.RB [ b | B | s | S | k | K | m | M | g | G ] +.SIZE_G \%ChunkSize .br Gives the size of chunk for snapshot, cache pool and thin pool logical volumes. Default unit is in kilobytes. .br -For \fIsnapshots\fP the value must be power of 2 between 4KiB and 512KiB +For snapshots the value must be power of 2 between 4KiB and 512KiB and the default value is 4KiB. .br -For \fIcache pools\fP the value must a multiple of 32KiB +For cache pools the value must a multiple of 32KiB between 32KiB and 1GiB. The default is 64KiB. .br -For \fIthin pools\fP the value must be a multiple of 64KiB +For thin pools the value must be a multiple of 64KiB between 64KiB and 1GiB. Default value starts with 64KiB and grows up to fit the pool metadata size within 128MiB, @@ -279,7 +297,7 @@ Default is \fBpassdown\fP. Configures thin pool behaviour when data space is exhausted. Default is \fBn\fPo. Device will queue I/O operations until target timeout -(see dm-thin-pool kernel module option \fIno_space_timeout\fP) +(see dm-thin-pool kernel module option \fPno_space_timeout\fP) expires. Thus configured system has a time to i.e. extend the size of thin pool data device. When set to \fBy\fPes, the I/O operation is immeditelly errored. @@ -313,7 +331,7 @@ PhysicalVolume(s) with the suffix \fB%PVS\fP, or (for a snapshot) as a percentage of the total space in the Origin Logical Volume with the suffix \fB%ORIGIN\fP (i.e. \fB100%ORIGIN\fP provides space for the whole origin). When expressed as a percentage, the number is treated -as an approximate upper limit for the total number of physical extents +as an approximate upper limit for the number of physical extents to be allocated (including extents used by any mirrors, for example). . .HP @@ -330,7 +348,7 @@ numbers are dynamically assigned. .BR \-\-metadataprofile .IR ProfileName .br -Uses and attaches the ProfileName configuration profile to the logical +Uses and attaches the \fIProfileName\fP configuration profile to the logical volume metadata. Whenever the logical volume is processed next time, the profile is automatically applied. If the volume group has another profile attached, the logical volume profile is preferred. @@ -448,7 +466,7 @@ Otherwise it is \fBn\fPo. . .HP .BR \-\-poolmetadatasize -.BR \fIMetadataVolumeSize [ b | B | s | S | k | K | m | M | g | G ] +.SIZE_G \%MetadataVolumeSize .br Sets the size of pool's metadata logical volume. Supported values are in range between 2MiB and 16GiB for thin pool, @@ -469,7 +487,7 @@ Default is \fBy\fPes. . .HP .BR \-\- [ raid ] maxrecoveryrate -.BR \fIRate [ b | B | s | S | k | K | m | M | g | G ] +.SIZE_G \%Rate .br Sets the maximum recovery rate for a RAID logical volume. \fIRate\fP is specified as an amount per second for each device in the array. @@ -478,7 +496,7 @@ recovery rate to 0 means it will be unbounded. . .HP .BR \-\- [ raid ] minrecoveryrate -.BR \fIRate [ b | B | s | S | k | K | m | M | g | G ] +.SIZE_G \%Rate .br Sets the minimum recovery rate for a RAID logical volume. \fIRate\fP is specified as an amount per second for each device in the array. @@ -498,7 +516,7 @@ a suitable value automatically. . .HP .BR \-R | \-\-regionsize -.BR \fIMirrorLogRegionSize [ b | B | s | S | k | K | m | M | g | G ] +.SIZE_G \%MirrorLogRegionSize .br A mirror is divided into regions of this size (in MiB), and the mirror log uses this granularity to track which regions are in sync. @@ -522,9 +540,7 @@ where the state of the flag is reported within \fBlv_attr\fP bits. . .HP .BR \-L | \-\-size -.BR \fILogicalVolumeSize [ b | B | s | S | k | K | m | M | \c -.BR g | G | t | T | p | P | e | E ] -.\" man2html cannot handle all those changes in 1 line +.SIZE_E \%LogicalVolumeSize .br Gives the size to allocate for the new logical volume. A size suffix of \fBB\fP for bytes, \fBS\fP for sectors as 512 bytes, @@ -533,8 +549,11 @@ A size suffix of \fBB\fP for bytes, \fBS\fP for sectors as 512 bytes, or \fBE\fP for exabytes is optional. .br Default unit is megabytes. -.TP -.IR \fB\-s ", " \fB\-\-snapshot " " OriginalLogicalVolume { Name | Path } +. +.HP +.BR \-s | \fB\-\-snapshot +.IR OriginalLogicalVolume { Name | Path } +.br Creates a snapshot logical volume (or snapshot) for an existing, so called original logical volume (or origin). Snapshots provide a 'frozen image' of the contents of the origin @@ -567,29 +586,35 @@ External origin volume can be used/shared for many thin volumes even from different thin pools. See .BR lvconvert (8) for online conversion to thin volumes with external origin. -.TP -.BR \-i ", " \-\-stripes " " \fIStripes +. +.HP +.BR \-i | \-\-stripes +.IR Stripes +.br Gives the number of stripes. This is equal to the number of physical volumes to scatter the logical volume. When creating a RAID 4/5/6 logical volume, the extra devices which are necessary for parity are -internally accounted for. Specifying -.BI \-i 3 +internally accounted for. Specifying \fB\-i 3\fP would use 3 devices for striped logical volumes, 4 devices for RAID 4/5, and 5 devices for RAID 6. Alternatively, RAID 4/5/6 will stripe across all PVs in the volume group or -all of the PVs specified if the -.B \-i +all of the PVs specified if the \fB\-i\fP argument is omitted. -.TP -.BR \-I ", " \-\-stripesize " " \fIStripeSize +. +.HP +.BR \-I | \-\-stripesize +.IR StripeSize +.br Gives the number of kilobytes for the granularity of the stripes. .br StripeSize must be 2^n (n = 2 to 9) for metadata in LVM1 format. For metadata in LVM2 format, the stripe size may be a larger power of 2 but must not exceed the physical extent size. -.TP -.IR \fB\-T ", " \fB\-\-thin +. +.HP +.BR \-T | \-\-thin +.br Creates thin pool or thin logical volume or both. Specifying the optional argument \fB\-\-size\fP or \fB\-\-extents\fP will cause the creation of the thin pool logical volume. @@ -600,81 +625,93 @@ thin pool and thin volume using this pool. See \fBlvmthin\fP(7) for more info about thin provisioning support. Thin provisioning requires device mapper kernel driver from kernel 3.2 or greater. -.TP -.IR \fB\-\-thinpool " " ThinPoolLogicalVolume { Name | Path } +. +.HP +.BR \-\-thinpool +.IR ThinPoolLogicalVolume { Name | Path } +.br Specifies the name of thin pool volume name. The other way to specify pool name is to append name to Volume group name argument. -.TP -.B \-\-type \fISegmentType +. +.HP +.BR \-\-type +.IR SegmentType +.br Creates a logical volume with the specified segment type. Supported types are: -.IR cache , -.IR cache-pool , -.IR error , -.IR linear , -.IR mirror, -.IR raid1 , -.IR raid4 , -.IR raid5_la , -.IR raid5_ls " (= " raid5 ), -.IR raid5_ra , -.IR raid5_rs , -.IR raid6_nc , -.IR raid6_nr , -.IR raid6_zr " (= " raid6 ) , -.IR raid10 , -.IR snapshot , -.IR striped, -.IR thin , -.IR thin-pool +.BR cache , +.BR cache-pool , +.BR error , +.BR linear , +.BR mirror, +.BR raid1 , +.BR raid4 , +.BR raid5_la , +.BR raid5_ls +.RB (= +.BR raid5 ), +.BR raid5_ra , +.BR raid5_rs , +.BR raid6_nc , +.BR raid6_nr , +.BR raid6_zr +.RB (= +.BR raid6 ), +.BR raid10 , +.BR snapshot , +.BR striped, +.BR thin , +.BR thin-pool or -.IR zero . +.BR zero . Segment type may have a commandline switch alias that will enable its use. When the type is not explicitly specified an implicit type is selected from combination of options: -.BR \-H | \-\-cache | \-\-cachepool " (" \fIcache -or -.IR cachepool ), -.BR \-T | \-\-thin | \-\-thinpool " (" \fIthin -or -.IR thinpool ), -.BR \-m | \-\-mirrors " (" \fIraid1 -or -.IR mirror ), -.BR \-s | \-\-snapshot | \-V | \-\-virtualsize " (" \fIsnapshot -or -.IR thin ), -.BR \-i | \-\-stripes " (" \fIstriped ). -Default type is \fIlinear\fP. -.TP -.BR \-V ", " \-\-virtualsize " " \fIVirtualSize [ \fIbBsSkKmMgGtTpPeE ] +.BR \-H | \-\-cache | \-\-cachepool +(cache or cachepool), +.BR \-T | \-\-thin | \-\-thinpool +(thin or thinpool), +.BR \-m | \-\-mirrors +(raid1 or mirror), +.BR \-s | \-\-snapshot | \-V | \-\-virtualsize +(snapshot or thin), +.BR \-i | \-\-stripes +(striped). +Default segment type is \fBlinear\fP. +. +.HP +.BR \-V | \-\-virtualsize +.SIZE_E \%VirtualSize +.br Creates a thinly provisioned device or a sparse device of the given size (in MiB by default). See .BR lvm.conf (5) -settings -.IR global / sparse_segtype_default +settings \fBglobal/sparse_segtype_default\fP to configure default sparse segment type. See \fBlvmthin\fP(7) for more info about thin provisioning support. Anything written to a sparse snapshot will be returned when reading from it. Reading from other areas of the device will return blocks of zeros. -Virtual snapshot is implemented by creating a hidden virtual device of the -requested size using the zero target. A suffix of _vorigin is used for -this device. Note: using sparse snapshots is not efficient for larger +Virtual snapshot (sparse snapshot) is implemented by creating +a hidden virtual device of the requested size using the zero target. +A suffix of _vorigin is used for this device. +Note: using sparse snapshots is not efficient for larger device sizes (GiB), thin provisioning should be used for this case. -.TP -.BR \-W ", " \-\-wipesignatures " {" \fIy | \fIn } +. +.HP +.BR \-W | \-\-wipesignatures +.RB { y | n } +.br Controls wiping of detected signatures on newly created Logical Volume. If this option is not specified, then by default signature wiping is done -each time the zeroing (\fB\-Z\fP/\fB\-\-zero\fP) is done. This default behaviour -can be controlled by -.IR allocation / wipe_signatures_when_zeroing_new_lvs +each time the zeroing ( +.BR \-Z | \-\-zero +) is done. This default behaviour +can be controlled by \fB\%allocation/wipe_signatures_when_zeroing_new_lvs\fP setting found in .BR lvm.conf (5). .br -If blkid wiping is used -.IR allocation / use_blkid_wiping -setting in +If blkid wiping is used \fBallocation/use_blkid_wiping\fP setting in .BR lvm.conf (5)) and LVM2 is compiled with blkid wiping support, then \fBblkid\fP(8) library is used to detect the signatures (use \fBblkid \-k\fP command to list the signatures that are recognized). @@ -682,17 +719,21 @@ Otherwise, native LVM2 code is used to detect signatures (MD RAID, swap and LUKS signatures are detected only in this case). .br Logical volume is not wiped if the read only flag is set. -.TP -.BR \-Z ", " \-\-zero " {" \fIy | \fIn } +. +.HP +.BR \-Z | \-\-zero +.RB { y | n } +.br Controls zeroing of the first 4KiB of data in the new logical volume. -Default is \fIy\fPes. +Default is \fBy\fPes. Snapshot COW volumes are always zeroed. Logical volume is not zeroed if the read only flag is set. - .br Warning: trying to mount an unzeroed logical volume can cause the system to hang. +. .SH Examples +. Creates a striped logical volume with 3 stripes, a stripe size of 8KiB and a size of 100MiB in the volume group named vg00. The logical volume name will be chosen by lvcreate: @@ -701,8 +742,8 @@ The logical volume name will be chosen by lvcreate: Creates a mirror logical volume with 2 sides with a useable size of 500 MiB. This operation would require 3 devices (or option -.BI \-\-alloc \ anywhere -) - two for the mirror devices and one for the disk log: +\fB\-\-alloc \%anywhere\fP) - two for the mirror +devices and one for the disk log: .sp .B lvcreate \-m1 \-L 500M vg00 @@ -794,8 +835,10 @@ volume (i.e. the origin LV), creating a cache LV. .\" Create a 1G cached LV "lvol1" with 10M cache pool "vg00/pool". .\" .sp .\" .B lvcreate \-\-cache \-L 1G \-n lv \-\-pooldatasize 10M vg00/pool - +. .SH SEE ALSO +. +.nh .BR lvm (8), .BR lvm.conf (5), .BR lvmcache (7), diff --git a/man/lvm.8.in b/man/lvm.8.in index 8ff31ae..31f7187 100644 --- a/man/lvm.8.in +++ b/man/lvm.8.in @@ -757,6 +757,7 @@ directly. . .SH SEE ALSO . +.nh .BR lvm.conf (5), .BR lvmcache (7), .BR lvmthin (7),

1 0

master - man: fsadm
by Zdenek Kabelac 06 Oct '15

06 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=4b1cadbd87a0814a… Commit: 4b1cadbd87a0814a46b36b07cfe94eda7baacb28 Parent: 2506275c3b5ec7db8c76ef7fdb07d4f3083a3684 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Mon Oct 5 12:28:00 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 6 15:21:36 2015 +0200 man: fsadm --- man/fsadm.8.in | 95 ++++++++++++++++++++++++++++++++++++----------------- scripts/fsadm.sh | 4 +- 2 files changed, 66 insertions(+), 33 deletions(-) diff --git a/man/fsadm.8.in b/man/fsadm.8.in index 2aecd65..02431c5 100644 --- a/man/fsadm.8.in +++ b/man/fsadm.8.in @@ -2,72 +2,105 @@ .SH "NAME" fsadm \(em utility to resize or check filesystem on a device .SH SYNOPSIS +. +.PD 0 +.ad l +.HP 5 .B fsadm .RI [ options ] -.B check -.I device -.sp +.BR check +.IR device +. +.HP .B fsadm .RI [ options ] -.B resize -.I device -.RI [ new_size [ BKMGTEP ]] -.sp +.BR resize +.IR device +.RI [ new_size ] +.PD +.ad b +. .SH DESCRIPTION +. fsadm utility checks or resizes the filesystem on a device. -It tries to use the same API for -.IR ext2 ", " ext3 ", " ext4 ", " ReiserFS " and " XFS +It tries to use the same API for +.BR ext2 , +.BR ext3 , +.BR ext4 , +.BR ReiserFS +.RB and +.BR XFS filesystem. +. .SH OPTIONS -.TP -.BR \-e ", " \-\-ext\-offline +. +.HP +.BR \-e | \-\-ext\-offline +.br Unmount ext2/ext3/ext4 filesystem before doing resize. -.TP -.BR \-f ", " \-\-force +. +.HP +.BR \-f | \-\-force +.br Bypass some sanity checks. -.TP -.BR \-h ", " \-\-help +. +.HP +.BR \-h | \-\-help +.br Display the help text. -.TP -.BR \-n ", " \-\-dry\-run +. +.HP +.BR \-n | \-\-dry\-run +.br Print commands without running them. -.TP -.BR \-v ", " \-\-verbose +. +.HP +.BR \-v | \-\-verbose +.br Be more verbose. -.TP -.BR \-y ", " \-\-yes +. +.HP +.BR \-y | \-\-yes +.br Answer "yes" at any prompts. -.TP -.I new_size +. +.HP +.BR \fInew_size [ B | K | M | G | T | P | E ] +.br Absolute number of filesystem blocks to be in the filesystem, or an absolute size using a suffix (in powers of 1024). If new_size is not supplied, the whole device is used. - +. .SH DIAGNOSTICS +. On successful completion, the status code is 0. A status code of 2 indicates the operation was interrupted by the user. A status code of 3 indicates the requested check operation could not be performed -because the filesystem is mounted and does not support an online +because the filesystem is mounted and does not support an online .BR fsck (8). A status code of 1 is used for other failures. - +. .SH EXAMPLES -Resize the filesystem on logical volume /dev/vg/test to 1000 megabytes. -If /dev/vg/test contains ext2/ext3/ext4 +. +Resize the filesystem on logical volume \fI/dev/vg/test\fP to 1000 megabytes. +If \fI/dev/vg/test\fP contains ext2/ext3/ext4 filesystem it will be unmounted prior the resize. All [y/n] questions will be answered 'y'. .sp .B fsadm \-e \-y resize /dev/vg/test 1000M +. .SH ENVIRONMENT VARIABLES +. .TP -.B TMPDIR -The temporary directory name for mount points. Defaults to "/tmp". +.B "TMPDIR " +The temporary directory name for mount points. Defaults to "\fI/tmp\fP". .TP .B DM_DEV_DIR The device directory name. -Defaults to "/dev" and must be an absolute path. +Defaults to "\fI/dev\fP" and must be an absolute path. .SH SEE ALSO +.nh .BR lvm (8), .BR lvresize (8), .BR lvm.conf (5), diff --git a/scripts/fsadm.sh b/scripts/fsadm.sh index f4ea796..e1aacd8 100755 --- a/scripts/fsadm.sh +++ b/scripts/fsadm.sh @@ -86,10 +86,10 @@ NL=' tool_usage() { echo "${TOOL}: Utility to resize or check the filesystem on a device" echo - echo " ${TOOL} [options] check device" + echo " ${TOOL} [options] check <device>" echo " - Check the filesystem on device using fsck" echo - echo " ${TOOL} [options] resize device [new_size[BKMGTPE]]" + echo " ${TOOL} [options] resize <device> [<new_size>[BKMGTPE]]" echo " - Change the size of the filesystem on device to new_size" echo echo " Options:"

1 0

master - cleanup: fix compiler warning
by Zdenek Kabelac 06 Oct '15

06 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=2506275c3b5ec7db… Commit: 2506275c3b5ec7db8c76ef7fdb07d4f3083a3684 Parent: 5f7a94a03e8ff61213844fd510e8af127475b501 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Sat Oct 3 22:10:31 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Tue Oct 6 14:57:30 2015 +0200 cleanup: fix compiler warning --- lib/cache/lvmetad.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/lib/cache/lvmetad.c b/lib/cache/lvmetad.c index db15866..20d2001 100644 --- a/lib/cache/lvmetad.c +++ b/lib/cache/lvmetad.c @@ -53,8 +53,8 @@ static int _log_debug_inequality(const char *name, struct dm_config_node *a, str log_debug_lvmetad("VG %s metadata inequality at %s / %s: %s / %s", name, a->key, b->key, av->v.str, bv->v.str); else if (a->v->type == DM_CFG_INT && b->v->type == DM_CFG_INT) - log_debug_lvmetad("VG %s metadata inequality at %s / %s: %li / %li", - name, a->key, b->key, (int64_t)av->v.i, (int64_t)bv->v.i); + log_debug_lvmetad("VG %s metadata inequality at %s / %s: " FMTi64 " / " FMTi64, + name, a->key, b->key, av->v.i, bv->v.i); else log_debug_lvmetad("VG %s metadata inequality at %s / %s: type %d / type %d", name, a->key, b->key, av->type, bv->type);

1 0

master - blkdeactivate: recognize and deactivate MD devices too
by Peter Rajnoha 06 Oct '15

06 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=5f7a94a03e8ff612… Commit: 5f7a94a03e8ff61213844fd510e8af127475b501 Parent: df59db6048a6be79acf231b98fd560b11cce7c7c Author: Peter Rajnoha <prajnoha(a)redhat.com> AuthorDate: Tue Oct 6 13:25:37 2015 +0200 Committer: Peter Rajnoha <prajnoha(a)redhat.com> CommitterDate: Tue Oct 6 13:25:41 2015 +0200 blkdeactivate: recognize and deactivate MD devices too --- WHATS_NEW_DM | 1 + man/blkdeactivate.8.in | 10 ++++++---- scripts/blkdeactivate.sh.in | 33 +++++++++++++++++++++++++++++++++ 3 files changed, 40 insertions(+), 4 deletions(-) diff --git a/WHATS_NEW_DM b/WHATS_NEW_DM index c1c48ea..e89ead0 100644 --- a/WHATS_NEW_DM +++ b/WHATS_NEW_DM @@ -1,5 +1,6 @@ Version 1.02.110 - ====================================== + Add support for recognition and deactivation of MD devices to blkdeactivate. Move target status functions out of libdm-deptree. Correct use of max_write_behind parameter when generating raid target line. Fix dm-event systemd service to make sure it is executed before mounting. diff --git a/man/blkdeactivate.8.in b/man/blkdeactivate.8.in index 990be4a..2eb4c6f 100644 --- a/man/blkdeactivate.8.in +++ b/man/blkdeactivate.8.in @@ -14,10 +14,11 @@ blkdeactivate \(em utility to deactivate block devices blkdeactivate utility deactivates block devices. If a device is mounted, the utility can unmount it automatically before trying to deactivate. The utility currently supports -device-mapper devices, including LVM volumes. -LVM volumes are handled directly using the \fBlvm\fP(8) command. -Other device-mapper based devices are handled using the -\fBdmsetup\fP(8) command. +device-mapper devices (DM), including LVM volumes and +software RAID MD devices. LVM volumes are handled directly +using the \fBlvm\fP(8) command, the rest of device-mapper +based devices are handled using the \fBdmsetup\fP(8) command. +MD devices are handled using the \fBmdadm\fP(8) command. .SH OPTIONS .TP .BR \-d ", " \-\-dmoption \ \fIdm_options\fP @@ -88,4 +89,5 @@ of device-mapper devices in case the deactivation fails and force removal. .BR dmsetup (8), .BR lsblk (8), .BR lvm (8), +.BR mdadm (8), .BR umount (8) diff --git a/scripts/blkdeactivate.sh.in b/scripts/blkdeactivate.sh.in index 993f151..dc1a006 100644 --- a/scripts/blkdeactivate.sh.in +++ b/scripts/blkdeactivate.sh.in @@ -37,6 +37,7 @@ SYS_BLK_DIR='/sys/block' UMOUNT="/bin/umount" DMSETUP="@sbindir@/dmsetup" LVM="@sbindir@/lvm" +MDADM="@sbindir@/mdadm" if $UMOUNT --help | grep -- "--all-targets" >$DEV_DIR/null; then UMOUNT_OPTS="--all-targets " @@ -47,6 +48,7 @@ else fi DMSETUP_OPTS="" LVM_OPTS="" +MDADM_OPTS="" LSBLK="/bin/lsblk -r --noheadings -o TYPE,KNAME,NAME,MOUNTPOINT" LSBLK_VARS="local devtype local kname local name local mnt" @@ -259,6 +261,28 @@ deactivate_lvm () { fi } +deactivate_md () { + local name=$(printf $name) + test -b "$DEV_DIR/$name" || return 0 + test -z ${SKIP_DEVICE_LIST["$kname"]} || return 1 + + # Skip MD device deactivation if MD tools missing. + test $MDADM_AVAILABLE -eq 0 && { + add_device_to_skip_list + return 1 + } + + deactivate_holders "$DEV_DIR/$name" || return 1 + + echo -n " [MD]: deactivating $devtype device $kname... " + if eval $MDADM $MDADM_OPTS -S "$name" $OUT $ERR; then + echo "done" + else + echo "skipping" + add_device_to_skip_list + fi +} + deactivate () { ###################################################################### # DEACTIVATION HOOKS FOR NEW DEVICE TYPES GO HERE! # @@ -277,6 +301,8 @@ deactivate () { deactivate_lvm elif test "${kname:0:3}" = "dm-"; then deactivate_dm + elif test "${kname:0:2}" = "md"; then + deactivate_md fi } @@ -392,6 +418,7 @@ set_env() { UMOUNT_OPTS+="-v" DMSETUP_OPTS+="-vvvv" LVM_OPTS+="-vvvv" + MDADM_OPTS+="-vv" else OUT="1>$DEV_DIR/null" fi @@ -401,6 +428,12 @@ set_env() { else LVM_AVAILABLE=0 fi + + if test -f $MDADM; then + MDADM_AVAILABLE=1 + else + MDADM_AVAILABLE=0 + fi } while test $# -ne 0; do

1 0

master - lockd: add error message for EEXIST
by David Teigland 02 Oct '15

02 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=df59db6048a6be79… Commit: df59db6048a6be79acf231b98fd560b11cce7c7c Parent: b33d7586e7f629818e881e26677f4431a47d50b5 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Fri Oct 2 15:41:23 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Fri Oct 2 15:41:23 2015 -0500 lockd: add error message for EEXIST The EEXIST error for LV locks is unusual, and was missing an explanatory error message. --- lib/locking/lvmlockd.c | 9 +++++++++ 1 files changed, 9 insertions(+), 0 deletions(-) diff --git a/lib/locking/lvmlockd.c b/lib/locking/lvmlockd.c index afbafcb..f741940 100644 --- a/lib/locking/lvmlockd.c +++ b/lib/locking/lvmlockd.c @@ -2010,6 +2010,15 @@ int lockd_lv_name(struct cmd_context *cmd, struct volume_group *vg, return 0; } + if (result == -EEXIST) { + /* + * This happens if lvchange tries to modify the LV with an ex + * LV lock when the LV is already active with a sh LV lock. + */ + log_error("LV is already locked with incompatible mode: %s/%s", vg->name, lv_name); + return 0; + } + if (result == -EMSGSIZE) { /* Another host probably extended lvmlock. */ if (!refreshed++) {

1 0

master - raid_manip: fix wrong image size allocation on raid10 "lvconvert --replace ..."
by Heinz Mauelshagen 02 Oct '15

02 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=b33d7586e7f62981… Commit: b33d7586e7f629818e881e26677f4431a47d50b5 Parent: fb957ef3223939424cf1faff4ca9dd104882ab46 Author: Heinz Mauelshagen <heinzm(a)redhat.com> AuthorDate: Fri Oct 2 17:09:37 2015 +0200 Committer: Heinz Mauelshagen <heinzm(a)redhat.com> CommitterDate: Fri Oct 2 17:09:37 2015 +0200 raid_manip: fix wrong image size allocation on raid10 "lvconvert --replace ..." --- lib/metadata/raid_manip.c | 9 ++++++--- 1 files changed, 6 insertions(+), 3 deletions(-) diff --git a/lib/metadata/raid_manip.c b/lib/metadata/raid_manip.c index 5919e3b..c420733 100644 --- a/lib/metadata/raid_manip.c +++ b/lib/metadata/raid_manip.c @@ -460,9 +460,12 @@ static int _alloc_image_components(struct logical_volume *lv, * individual devies, we must specify how large the individual device * is along with the number we want ('count'). */ - extents = (segtype->parity_devs) ? - (lv->le_count / (seg->area_count - segtype->parity_devs)) : - lv->le_count; + if (segtype_is_raid10(segtype)) + extents = lv->le_count / (seg->area_count / 2); /* we enforce 2 mirrors right now */ + else + extents = (segtype->parity_devs) ? + (lv->le_count / (seg->area_count - segtype->parity_devs)) : + lv->le_count; if (!(ah = allocate_extents(lv->vg, NULL, segtype, 0, count, count, region_size, extents, pvs,

1 0

master - raid: Add metadata dev information to reports.
by Alasdair Kergon 02 Oct '15

02 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=fb957ef322393942… Commit: fb957ef3223939424cf1faff4ca9dd104882ab46 Parent: 5e5d48348b0c7afdd7a3fb56ddd6875a8b81dc68 Author: Alasdair G Kergon <agk(a)redhat.com> AuthorDate: Fri Oct 2 10:09:28 2015 +0100 Committer: Alasdair G Kergon <agk(a)redhat.com> CommitterDate: Fri Oct 2 10:09:28 2015 +0100 raid: Add metadata dev information to reports. Add metadata_devices and seg_metadata_le_ranges report fields. Currently only defined for raid, but should probably be extended to all other segment types that don't report all their device usage in the 'devices' field. --- WHATS_NEW | 1 + lib/metadata/lv.c | 31 ++++++++++++++++++++++++------- lib/metadata/lv.h | 2 ++ lib/report/columns.h | 2 ++ lib/report/properties.c | 5 +++++ lib/report/report.c | 24 ++++++++++++++++++++++++ 6 files changed, 58 insertions(+), 7 deletions(-) diff --git a/WHATS_NEW b/WHATS_NEW index d693ab8..5da1552 100644 --- a/WHATS_NEW +++ b/WHATS_NEW @@ -1,5 +1,6 @@ Version 2.02.133 - ====================================== + Add metadata_devices and seg_metadata_le_ranges report fields for raid vols. Fix lvm2-{activation,clvmd,cmirrord,monitor} service to exec before mounting. Version 2.02.132 - 22nd September 2015 diff --git a/lib/metadata/lv.c b/lib/metadata/lv.c index e0080b1..3126086 100644 --- a/lib/metadata/lv.c +++ b/lib/metadata/lv.c @@ -29,7 +29,7 @@ static struct utsname _utsname; static int _utsinit = 0; static char *_format_pvsegs(struct dm_pool *mem, const struct lv_segment *seg, - int range_format) + int range_format, int metadata_areas_only) { unsigned int s; const char *name = NULL; @@ -41,13 +41,19 @@ static char *_format_pvsegs(struct dm_pool *mem, const struct lv_segment *seg, return NULL; } + if (metadata_areas_only && (!seg_is_raid(seg) || lv_is_raid_metadata(seg->lv) || lv_is_raid_image(seg->lv))) + goto out; + for (s = 0; s < seg->area_count; s++) { - switch (seg_type(seg, s)) { + switch (metadata_areas_only ? seg_metatype(seg, s) : seg_type(seg, s)) { case AREA_LV: - name = seg_lv(seg, s)->name; - extent = seg_le(seg, s); + name = metadata_areas_only ? seg_metalv(seg, s)->name : seg_lv(seg, s)->name; + extent = metadata_areas_only ? seg_le(seg, s) : 0; break; case AREA_PV: + /* Raid metadata never uses PVs directly */ + if (metadata_areas_only) + continue; name = dev_name(seg_dev(seg, s)); extent = seg_pe(seg, s); break; @@ -79,7 +85,7 @@ static char *_format_pvsegs(struct dm_pool *mem, const struct lv_segment *seg, if (range_format) { if (dm_snprintf(extent_str, sizeof(extent_str), - FMTu32, extent + seg->area_len - 1) < 0) { + FMTu32, metadata_areas_only ? extent + seg_metalv(seg, s)->le_count - 1 : extent + seg->area_len - 1) < 0) { log_error("Extent number dm_snprintf failed"); return NULL; } @@ -96,6 +102,7 @@ static char *_format_pvsegs(struct dm_pool *mem, const struct lv_segment *seg, } } +out: if (!dm_pool_grow_object(mem, "\0", 1)) { log_error("dm_pool_grow_object failed"); return NULL; @@ -106,12 +113,22 @@ static char *_format_pvsegs(struct dm_pool *mem, const struct lv_segment *seg, char *lvseg_devices(struct dm_pool *mem, const struct lv_segment *seg) { - return _format_pvsegs(mem, seg, 0); + return _format_pvsegs(mem, seg, 0, 0); +} + +char *lvseg_metadata_devices(struct dm_pool *mem, const struct lv_segment *seg) +{ + return _format_pvsegs(mem, seg, 0, 1); } char *lvseg_seg_pe_ranges(struct dm_pool *mem, const struct lv_segment *seg) { - return _format_pvsegs(mem, seg, 1); + return _format_pvsegs(mem, seg, 1, 0); +} + +char *lvseg_seg_metadata_le_ranges(struct dm_pool *mem, const struct lv_segment *seg) +{ + return _format_pvsegs(mem, seg, 1, 1); } char *lvseg_tags_dup(const struct lv_segment *seg) diff --git a/lib/metadata/lv.h b/lib/metadata/lv.h index a839489..172d29f 100644 --- a/lib/metadata/lv.h +++ b/lib/metadata/lv.h @@ -101,7 +101,9 @@ char *lvseg_cachemode_dup(struct dm_pool *mem, const struct lv_segment *seg); char *lvseg_monitor_dup(struct dm_pool *mem, const struct lv_segment *seg); char *lvseg_tags_dup(const struct lv_segment *seg); char *lvseg_devices(struct dm_pool *mem, const struct lv_segment *seg); +char *lvseg_metadata_devices(struct dm_pool *mem, const struct lv_segment *seg); char *lvseg_seg_pe_ranges(struct dm_pool *mem, const struct lv_segment *seg); +char *lvseg_seg_metadata_le_ranges(struct dm_pool *mem, const struct lv_segment *seg); char *lv_time_dup(struct dm_pool *mem, const struct logical_volume *lv, int iso_mode); char *lv_host_dup(struct dm_pool *mem, const struct logical_volume *lv); int lv_set_creation(struct logical_volume *lv, diff --git a/lib/report/columns.h b/lib/report/columns.h index 3b58aee..6041eeb 100644 --- a/lib/report/columns.h +++ b/lib/report/columns.h @@ -190,7 +190,9 @@ FIELD(SEGS, seg, SIZ, "SSize", list, 5, segsize, seg_size, "Size of segment in c FIELD(SEGS, seg, SIZ, "SSize", list, 5, segsizepe, seg_size_pe, "Size of segment in physical extents.", 0) FIELD(SEGS, seg, STR_LIST, "Seg Tags", tags, 8, tags, seg_tags, "Tags, if any.", 0) FIELD(SEGS, seg, STR, "PE Ranges", list, 9, peranges, seg_pe_ranges, "Ranges of Physical Extents of underlying devices in command line format.", 0) +FIELD(SEGS, seg, STR, "Metadata LE Ranges", list, 18, metadataleranges, seg_metadata_le_ranges, "Ranges of Logical Extents of underlying metadata devices in command line format.", 0) FIELD(SEGS, seg, STR, "Devices", list, 7, devices, devices, "Underlying devices used with starting extent numbers.", 0) +FIELD(SEGS, seg, STR, "Metadata Devs", list, 13, metadatadevices, metadata_devices, "Underlying metadata devices used with starting extent numbers.", 0) FIELD(SEGS, seg, STR, "Monitor", list, 7, segmonitor, seg_monitor, "Dmeventd monitoring status of the segment.", 0) FIELD(SEGS, seg, STR, "Cache Policy", list, 12, cache_policy, cache_policy, "The cache policy (cached segments only).", 0) FIELD(SEGS, seg, STR_LIST, "Cache Settings", list, 14, cache_settings, cache_settings, "Cache settings/parameters (cached segments only).", 0) diff --git a/lib/report/properties.c b/lib/report/properties.c index 72dcaff..262cc26 100644 --- a/lib/report/properties.c +++ b/lib/report/properties.c @@ -463,8 +463,13 @@ GET_LVSEG_STR_PROPERTY_FN(seg_tags, lvseg_tags_dup(lvseg)) GET_LVSEG_STR_PROPERTY_FN(seg_pe_ranges, lvseg_seg_pe_ranges(lvseg->lv->vg->vgmem, lvseg)) #define _seg_pe_ranges_set prop_not_implemented_set +GET_LVSEG_STR_PROPERTY_FN(seg_metadata_le_ranges, + lvseg_seg_metadata_le_ranges(lvseg->lv->vg->vgmem, lvseg)) +#define _seg_metadata_le_ranges_set prop_not_implemented_set GET_LVSEG_STR_PROPERTY_FN(devices, lvseg_devices(lvseg->lv->vg->vgmem, lvseg)) #define _devices_set prop_not_implemented_set +GET_LVSEG_STR_PROPERTY_FN(metadata_devices, lvseg_metadata_devices(lvseg->lv->vg->vgmem, lvseg)) +#define _metadata_devices_set prop_not_implemented_set GET_LVSEG_STR_PROPERTY_FN(seg_monitor, lvseg_monitor_dup(lvseg->lv->vg->vgmem, lvseg)) #define _seg_monitor_set prop_not_implemented_set diff --git a/lib/report/report.c b/lib/report/report.c index bf2f1c8..7ac1ac5 100644 --- a/lib/report/report.c +++ b/lib/report/report.c @@ -1304,6 +1304,18 @@ static int _devices_disp(struct dm_report *rh __attribute__((unused)), struct dm return _field_set_value(field, str, NULL); } +static int _metadatadevices_disp(struct dm_report *rh __attribute__((unused)), struct dm_pool *mem, + struct dm_report_field *field, + const void *data, void *private __attribute__((unused))) +{ + char *str; + + if (!(str = lvseg_metadata_devices(mem, (const struct lv_segment *) data))) + return_0; + + return _field_set_value(field, str, NULL); +} + static int _peranges_disp(struct dm_report *rh __attribute__((unused)), struct dm_pool *mem, struct dm_report_field *field, const void *data, void *private __attribute__((unused))) @@ -1316,6 +1328,18 @@ static int _peranges_disp(struct dm_report *rh __attribute__((unused)), struct d return _field_set_value(field, str, NULL); } +static int _metadataleranges_disp(struct dm_report *rh __attribute__((unused)), struct dm_pool *mem, + struct dm_report_field *field, + const void *data, void *private __attribute__((unused))) +{ + char *str; + + if (!(str = lvseg_seg_metadata_le_ranges(mem, (const struct lv_segment *) data))) + return_0; + + return _field_set_value(field, str, NULL); +} + static int _tags_disp(struct dm_report *rh, struct dm_pool *mem, struct dm_report_field *field, const void *data, void *private)

1 0

master - man lvmcache: updates for mode and policy
by David Teigland 01 Oct '15

01 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=5e5d48348b0c7afd… Commit: 5e5d48348b0c7afdd7a3fb56ddd6875a8b81dc68 Parent: 26da6a3e10d2b8a84326288b53c8e9a6d82a3846 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Thu Oct 1 11:39:07 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Thu Oct 1 14:23:39 2015 -0500 man lvmcache: updates for mode and policy Correct some things, e.g. set mode and policy on the cache lv, not the pool, lvm.conf field for mode changed. Add smq which was missing. Make the sections on cache mode and cache policy consistent in structure and style. --- man/lvmcache.7.in | 74 ++++++++++++++++++++++++++++++----------------------- 1 files changed, 42 insertions(+), 32 deletions(-) diff --git a/man/lvmcache.7.in b/man/lvmcache.7.in index 7257386..44709d1 100644 --- a/man/lvmcache.7.in +++ b/man/lvmcache.7.in @@ -195,6 +195,7 @@ Users who are concerned about the possibility of failures in their fast devices that could lead to data loss might consider making their cache pool sub-LVs redundant. +.I Example .nf 0. Create an origin LV we wish to cache # lvcreate \-L 10G \-n lv1 vg /dev/slow_devs @@ -228,14 +229,18 @@ from the cache pool back to the origin LV. This mode will increase performance, but the loss of a device associated with the cache pool LV can result in lost data. -The cache mode can be specified with the --cachemode option when a cache -pool LV is created. +With the \-\-cachemode option, the cache mode can be set when creating a +cache LV, or changed on an existing cache LV. The current cache mode of a +cache LV can be displayed with the cache_mode reporting option: + +.B lvs \-o+cache_mode VG/CacheLV .BR lvm.conf (5) -.B cache_pool_cachemode +.B allocation/cache_mode .br defines the default cache mode. +.I Example .nf 0. Create an origin LV we wish to cache (yours may already exist) # lvcreate \-L 10G \-n lv1 vg /dev/slow @@ -246,50 +251,55 @@ defines the default cache mode. 2. Create a cache metadata LV # lvcreate \-L 8M \-n cache1meta vg /dev/fast -3. Create a cache pool LV specifying cache mode "writethrough" -# lvconvert \-\-type cache\-pool \-\-poolmetadata vg/cache1meta \\ - \-\-cachemode writethrough vg/cache1 +3. Create a cache pool LV +# lvconvert \-\-type cache\-pool \-\-poolmetadata vg/cache1meta vg/cache1 -4. Create a cache LV by combining the cache pool LV and origin LV -# lvconvert \-\-type cache \-\-cachepool vg/cache1 vg/lv1 +4. Create a cache LV by combining the cache pool LV and origin LV, + and use the writethrough cache mode. +# lvconvert \-\-type cache \-\-cachepool vg/cache1 \\ + \-\-cachemode writethrough vg/lv1 .fi -The cache mode can be changed on an existing LV with the command: -.B lvconvert --cachemode writethrough|writeback VG/CacheLV +.SS Cache policy +\& -.SS Cache policy & policy settings +The cache subsystem has additional per-LV parameters: the cache policy to +use, and possibly tunable parameters for the cache policy. Three policies +are currently available: "smq" is the default policy, "mq" is an older +implementation, and "cleaner" is used to force the cache to write back +(flush) all cached writes to the origin LV. -\& +The "mq" policy has a number of tunable parameters. The defaults are +chosen to be suitable for the majority of systems, but in special +circumstances, changing the settings can improve performance. -The cache subsystem has an additional per-LV parameter, namely the cache policy -to use, and possibly the tunable parameters of the said cache policy. In the -current implementation, two policies are available, "mq" which is the default -policy and "cleaner" which is used to force the cache to write back (flush) all -cached writes to the origin LV. Moreover, the "mq" policy has a number of -tunable parameters: the defaults are chosen to be suitable for the vast -majority of systems. However, under special circumstances, changing the tunable -settings of the cache policy can improve performance. +With the \-\-cachepolicy and \-\-cachesettings options, the cache policy +and settings can be set when creating a cache LV, or changed on an +existing cache LV (both options can be used together). The current cache +policy and settings of a cache LV can be displayed with the cache_policy +and cache_settings reporting options: -On an existing cache LV, the policy can be set (to "mq") and the cache settings -can be changed using commands like these: +.B lvs \-o+cache_policy,cache_settings VG/CacheLV .I Example - .nf -# lvchange \-\-cachepolicy mq vg/lv1 -# lvchange \-\-cachesettings \(aqmigration_threshold=2048 random_threshold=4\(aq \\ - vg/lv1 +Change the cache policy and settings of an existing cache LV. +# lvchange \-\-cachepolicy mq \-\-cachesettings \\ + \(aqmigration_threshold=2048 random_threshold=4\(aq vg/lv1 .fi -Both commands can be combined, setting both cache policy and its settings -together. Moreover, when creating a cache LV for the first time (using -lvcreate), the \-\-cachepolicy and \-\-cachesettings parameters can be used as -well. The current policy and the policy settings can be listed using the lvs -command, using 'cache_policy' and 'cache_settings' fields: +.BR lvm.conf (5) +.B allocation/cache_policy +.br +defines the default cache policy. + +.BR lvm.conf (5) +.B allocation/cache_settings +.br +defines the default cache settings. -# lvs -o +cache_policy,cache_settings .SS Spare metadata LV

1 0

master - config: improve cache_settings description
by David Teigland 01 Oct '15

01 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=26da6a3e10d2b8a8… Commit: 26da6a3e10d2b8a84326288b53c8e9a6d82a3846 Parent: 4c2cc782aa8c7e6bdc18339215f65e82ea1489ae Author: David Teigland <teigland(a)redhat.com> AuthorDate: Thu Oct 1 11:45:07 2015 -0500 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Thu Oct 1 14:23:39 2015 -0500 config: improve cache_settings description --- lib/config/config_settings.h | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/lib/config/config_settings.h b/lib/config/config_settings.h index 9017043..9cc444f 100644 --- a/lib/config/config_settings.h +++ b/lib/config/config_settings.h @@ -474,8 +474,8 @@ cfg(allocation_cache_policy_CFG, "cache_policy", allocation_CFG_SECTION, CFG_PRO "otherwise the older mq (Multiqueue) policy is selected.\n") cfg_section(allocation_cache_settings_CFG_SECTION, "cache_settings", allocation_CFG_SECTION, CFG_PROFILABLE | CFG_DEFAULT_COMMENTED, vsn(2, 2, 128), 0, NULL, - "Individual settings for policies.\n" - "See the help for individual policies for more info.\n") + "Settings for the cache policy.\n" + "See documentation for individual cache policies for more info.\n") cfg_section(policy_settings_CFG_SUBSECTION, "policy_settings", allocation_cache_settings_CFG_SECTION, CFG_NAME_VARIABLE | CFG_SECTION_NO_CHECK | CFG_PROFILABLE | CFG_DEFAULT_COMMENTED, vsn(2, 2, 128), 0, NULL, "Replace this subsection name with a policy name.\n"

1 0

master - tests: ensure test will not block
by Zdenek Kabelac 01 Oct '15

01 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=4c2cc782aa8c7e6b… Commit: 4c2cc782aa8c7e6bdc18339215f65e82ea1489ae Parent: 05e7fdd5ce3eb5db827aa8b8ab9e07571d9a857a Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 1 16:45:14 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 1 16:45:14 2015 +0200 tests: ensure test will not block This test is somewhat 'stupid' and simulates mostly unsupported situation - so avoid blocking here. --- test/shell/snapshot-remove-dmsetup.sh | 10 +++++++++- 1 files changed, 9 insertions(+), 1 deletions(-) diff --git a/test/shell/snapshot-remove-dmsetup.sh b/test/shell/snapshot-remove-dmsetup.sh index 610d3c2..952a7bb 100644 --- a/test/shell/snapshot-remove-dmsetup.sh +++ b/test/shell/snapshot-remove-dmsetup.sh @@ -58,9 +58,17 @@ dmsetup resume $vg1-snap # Try how force removal works dmsetup suspend $vg-$lv1 # needs to fail as device is still open -not dmsetup remove --force $vg1-snap +not dmsetup remove --force $vg1-snap & + +sleep .5 + +dmsetup table $vg1-snap | tee out +should grep -i error out + dmsetup resume $vg-$lv1 +wait + # check it really is now 'error' target dmsetup table $vg1-snap | tee out grep error out

1 0

master - tests: use proper target name for check
by Zdenek Kabelac 01 Oct '15

01 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=05e7fdd5ce3eb5db… Commit: 05e7fdd5ce3eb5db827aa8b8ab9e07571d9a857a Parent: 796e3fb7e4dc9d996316b0fcc4ed09a2cc800ade Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 1 16:05:51 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 1 16:05:51 2015 +0200 tests: use proper target name for check --- test/shell/snapshot-merge.sh | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/test/shell/snapshot-merge.sh b/test/shell/snapshot-merge.sh index ff4f8d1..f2ad2cb 100644 --- a/test/shell/snapshot-merge.sh +++ b/test/shell/snapshot-merge.sh @@ -12,7 +12,7 @@ . lib/inittest # test if snapshot-merge target is available -aux target_at_least snapshot-merge 1 0 0 || skip +aux target_at_least dm-snapshot-merge 1 0 0 || skip which mkfs.ext3 || skip

1 0

master - man: update lvm pages
by Zdenek Kabelac 01 Oct '15

01 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=796e3fb7e4dc9d99… Commit: 796e3fb7e4dc9d996316b0fcc4ed09a2cc800ade Parent: 867a36b4197e2abe6efd3b16e229590eedc566a3 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Wed Sep 23 11:28:54 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 1 15:03:34 2015 +0200 man: update lvm pages --- man/clvmd.8.in | 141 +++++++---- man/lvchange.8.in | 344 ++++++++++++++++---------- man/lvcreate.8.in | 418 ++++++++++++++++++------------ man/lvm.8.in | 734 +++++++++++++++++++++++++++++++++-------------------- 4 files changed, 1017 insertions(+), 620 deletions(-) diff --git a/man/clvmd.8.in b/man/clvmd.8.in index 90de980..9b9310d 100644 --- a/man/clvmd.8.in +++ b/man/clvmd.8.in @@ -1,125 +1,159 @@ .TH CLVMD 8 "LVM TOOLS #VERSION#" "Red Hat Inc" \" -*- nroff -*- +. .SH NAME +. clvmd \(em cluster LVM daemon +. .SH SYNOPSIS +. +.ad l .B clvmd .RB [ \-C ] .RB [ \-d .RI [ value ]] .RB [ \-E -.IR "lock uuid" ] +.IR lock_uuid ] .RB [ \-f ] .RB [ \-h ] .RB [ \-I -.IR "cluster manager" ] +.IR cluster_manager ] .RB [ \-R ] .RB [ \-S ] .RB [ \-t .IR timeout ] .RB [ \-T -.IR "start timeout" ] +.IR start_timeout ] .RB [ \-V ] +.ad b +. .SH DESCRIPTION +. clvmd is the daemon that distributes LVM metadata updates around a cluster. It must be running on all nodes in the cluster and will give an error if a node in the cluster does not have this daemon running. +. .SH OPTIONS -.TP -.BR \-d \ [ \fIvalue ] +. +.HP +.BR \-C +.br +Only valid if \fB\-d\fP is also specified. +Tells all clvmds in a cluster to enable/disable debug logging. +Without this switch, only the local clvmd will change its debug level to that +given with \fB\-d\fP. +.br +This does not work correctly if specified on the command-line that starts clvmd. +If you want to start clvmd \fBand\fP +enable cluster-wide logging then the command needs to be issued twice, eg: +.br +.BR clvmd +.br +.BR clvmd\ \-d2 +. +.HP +.BR \-d +.RI [ value ] +.br Set debug logging level. If \fB\-d\fP is specified without a \fIvalue\fP then 1 is assumed. \fIValue\fP can be: .PD 0 .IP -.B 0 +.BR 0 \(em Disabled .IP -.B 1 +.BR 1 \(em Sends debug logs to stderr (implies \fB\-f\fP) .IP -.B 2 +.BR 2 \(em Sends debug logs to \fBsyslog\fP(3) .PD -.TP -.B \-C -Only valid if -.B \-d -is also specified. Tells all clvmds in a cluster to enable/disable debug logging. -Without this switch, only the local clvmd will change its debug level to that -given with -.B \-d . +.HP +.BR \-E +.IR lock_uuid .br -This does not work correctly if specified on the command-line that starts clvmd. -If you want to start clvmd -.B and -enable cluster-wide logging then the command needs to be issued twice, eg: -.br -.B clvmd -.br -.B clvmd \-d2 +Pass lock uuid to be reacquired exclusively when clvmd is restarted. +. +.HP +.BR \-f .br -.TP -.BR \-E "\fI lock uuid" -Pass \fIlock uuid\fP to be reacquired exclusively when clvmd is restarted. -.TP -.B \-f Don't fork, run in the foreground. -.TP -.B \-h +. +.HP +.BR \-h +.br Show help information. -.TP -.BR \-I "\fI cluster manager" -Selects the \fIcluster manager\fP to use for locking and internal +. +.HP +.BR \-I +.IR cluster_manager +.br +Selects the cluster manager to use for locking and internal communications. As it is quite possible to have multiple managers available on the same system you might have to manually specify this option to override the search. -By default, omit \fB-I\fP is equivalent to \fB\-Iauto\fP. \fBClvmd\fP -will use the first cluster manager that succeeds, and it checks them in a -predefined order cman,corosync,openais. +By default, omit \fB-I\fP is equivalent to \fB\-Iauto\fP. +Clvmd will use the first cluster manager that succeeds, +and it checks them in a predefined order +.BR cman , +.BR corosync , +.BR openais . The available managers will be listed by order as part of the \fBclvmd \-h\fP output. -.TP -.B \-R +. +.HP +.BR \-R +.br Tells all the running instance of \fBclvmd\fP in the cluster to reload their device cache and re-read the lvm configuration file \fBlvm.conf\fP(5). This command should be run whenever the devices on a cluster system are changed. -.TP -.B \-S +. +.HP +.BR \-S +.br Tells the running \fBclvmd\fP to exit and reexecute itself, for example at the end of a package upgrade. The new instance is instructed to reacquire any locks in the same state as they were previously held. (Alternative methods of restarting the daemon have the side effect of changing exclusive LV locks into shared locks.) -.TP -.BR \-t "\fI timeout" +. +.HP +.BR \-t +.IR timeout +.br Specifies the \fItimeout\fP for commands to run around the cluster. This should not be so small that commands with many disk updates to do will fail, so you may need to increase this on systems with very large disk farms. The default is 60 seconds. -.TP -.BR \-T "\fI start timeout" -Specifies the \fIstart timeout\fP for \fBclvmd\fP daemon startup. If the +. +.HP +.BR \-T +.IR start_timeout +.br +Specifies the start timeout for \fBclvmd\fP daemon startup. If the daemon does not report that it has started up within this time then the parent command will exit with status of 5. This does NOT mean that \fBclvmd\fP has not started! What it means is that the startup has been delayed for some reason; the most likely cause of this is an inquorate cluster though it could be due to locking latencies on a cluster with large numbers of logical volumes. If you get the return code of 5 it is usually not necessary to -restart \fBclvmd\fP — it will start as soon as that blockage has cleared. +restart \fBclvmd\fP it will start as soon as that blockage has cleared. This flag is to allow startup scripts to exit in a timely fashion even if the cluster is stalled for some reason. -The default is 0 (no timeout) and the value is in seconds. Don't set this too +The default is \fB0\fP (no timeout) and the value is in seconds. Don't set this too small or you will experience spurious errors. 10 or 20 seconds might be sensible. This timeout will be ignored if you start \fBclvmd\fP with the \fB\-d\fP. -.TP -.B \-V +. +.HP +.BR \-V +.br Display the version of the cluster LVM daemon. - +. .SH ENVIRONMENT VARIABLES .TP .B LVM_CLVMD_BINARY @@ -129,7 +163,10 @@ Defaults to \fI#CLVMD_PATH#\fP. .B LVM_BINARY The LVM2 binary to use. Defaults to \fI#LVM_PATH#\fP. - +.SH FILES +.I #CLVMD_PATH# +.br +.I #LVM_PATH# .SH SEE ALSO .BR syslog (3), .BR lvm.conf (5), diff --git a/man/lvchange.8.in b/man/lvchange.8.in index e8babba..62bdca1 100644 --- a/man/lvchange.8.in +++ b/man/lvchange.8.in @@ -1,21 +1,30 @@ .TH LVCHANGE 8 "LVM TOOLS #VERSION#" "Sistina Software UK" \" -*- nroff -*- + +.de UNITS +.. + +. .SH NAME +. lvchange \(em change attributes of a logical volume +. .SH SYNOPSIS +. +.ad l .B lvchange -.RB [ \-\-addtag -.IR Tag ] -.RB [ \-A | \-\-autobackup -.RB { y | n }] .RB [ \-a | \-\-activate .RB [ a ][ e | s | l ]{ y | n }] .RB [ \-\-activationmode .RB { complete | degraded | partial }] +.RB [ \-\-addtag +.IR Tag ] +.RB [ \-K | \-\-ignoreactivationskip ] .RB [ \-k | \-\-setactivationskip .RB { y | n }] -.RB [ \-K | \-\-ignoreactivationskip ] .RB [ \-\-alloc .IR AllocationPolicy ] +.RB [ \-A | \-\-autobackup +.RB { y | n }] .RB [ \-\-cachepolicy .IR policy ] .RB [ \-\-cachesettings @@ -33,13 +42,24 @@ lvchange \(em change attributes of a logical volume .RB { ignore | nopassdown | passdown }] .RB [ \-\-errorwhenfull .RB { y | n }] -.RB [ \-\-resync ] .RB [ \-h | \-? | \-\-help ] -.RB [ \-\-ignorelockingfailure ] -.RB [ \-\-ignoremonitoring ] -.RB [ \-\-ignoreskippedcluster ] +.RB \%[ \-\-ignorelockingfailure ] +.RB \%[ \-\-ignoremonitoring ] +.RB \%[ \-\-ignoreskippedcluster ] +.RB \%[ \-\-metadataprofile +.IR ProfileName ] .RB [ \-\-monitor .RB { y | n }] +.RB [ \-\-noudevsync ] +.RB [ \-P | \-\-partial ] +.RB [ \-p | \-\-permission +.RB { r | rw }] +.RB [ \-M | \-\-persistent +.RB { y | n } +.RB [ \-\-major +.IR major ] +.RB [ \-\-minor +.IR minor ]] .RB [ \-\-poll .RB { y | n }] .RB [ \-\- [ raid ] maxrecoveryrate @@ -52,41 +72,38 @@ lvchange \(em change attributes of a logical volume .IR IOCount ] .RB [ \-\- [ raid ] writemostly .BR \fIPhysicalVolume [ : { y | n | t }]] -.RB [ \-\-sysinit ] -.RB [ \-\-noudevsync ] -.RB [ \-\-metadataprofile -.IR ProfileName ] -.RB [ \-M | \-\-persistent -.RB { y | n } -.RB [ \-\-minor -.IR minor ] -.RB [ \-\-major -.IR major ]] -.RB [ \-P | \-\-partial ] -.RB [ \-p | \-\-permission -.RB { r | rw }] .RB [ \-r | \-\-readahead .RB { \fIReadAheadSectors | auto | none }] .RB [ \-\-refresh ] +.RB [ \-\-resync ] .RB [ \-S | \-\-select .IR Selection ] +.RB [ \-\-sysinit ] .RB [ \-t | \-\-test ] .RB [ \-v | \-\-verbose ] .RB [ \-Z | \-\-zero .RB { y | n }] .RI [ LogicalVolumePath ...] +.ad b +. .SH DESCRIPTION +. lvchange allows you to change the attributes of a logical volume including making them known to the kernel ready for use. +. .SH OPTIONS +. See \fBlvm\fP(8) for common options. -.TP -.BR \-a ", " \-\-activate " [" a ][ e | s | l ]{ y | n } +. +.HP +.BR \-a | \-\-activate +.RB [ a ][ e | s | l ]{ y | n } +.br Controls the availability of the logical volumes for use. Communicates with the kernel device-mapper driver via libdevmapper to activate (\fB\-ay\fP) or deactivate (\fB\-an\fP) the logical volumes. -.IP +.br Activation of a logical volume creates a symbolic link \fI/dev/VolumeGroupName/LogicalVolumeName\fP pointing to the device node. This link is removed on deactivation. @@ -95,19 +112,16 @@ this symbolic link and present this as the name of the device. The location and name of the underlying device node may depend on the distribution and configuration (e.g. udev) and might change from release to release. -.IP +.br If autoactivation option is used (\fB\-aay\fP), the logical volume is activated only if it matches an item in -the -.B activation/auto_activation_volume_list +the \fBactivation/auto_activation_volume_list\fP set in \fBlvm.conf\fP(5). If this list is not set, then all volumes are considered for activation. The \fB\-aay\fP option should be also used during system boot so it's possible to select which volumes to activate using -the -.B activation/auto_activation_volume_list -setting. -.IP +the \fBactivation/auto_activation_volume_list\fP setting. +.br In a clustered VG, clvmd is used for activation, and the following options are possible: @@ -137,9 +151,11 @@ LVs with snapshots are always activated exclusively because they can only be used on one node at once. For local VGs \fB\-ay\fP, \fB\-aey\fP, and \fB\-asy\fP are all equivalent. - -.TP -.BR \-\-activationmode " {" complete | degraded | partial } +. +.HP +.BR \-\-activationmode +.RB { complete | degraded | partial } +.br The activation mode determines whether logical volumes are allowed to activate when there are physical volumes missing (e.g. due to a device failure). \fBcomplete\fP is the most restrictive; allowing only those @@ -150,106 +166,168 @@ considered a RAID logical volume. The "\fIraid1\fP" segment type should be used instead.) Finally, \fBpartial\fP allows any logical volume to be activated even if portions are missing due to a missing or failed PV. This last option should only be used when performing recovery or -repair operations. \fBdegraded\fP is the default mode. To change it, modify -.B activation_mode -in -.BR lvm.conf (5). -.TP -.BR \-k ", " \-\-setactivationskip " {" y | n } +repair operations. \fBdegraded\fP is the default mode. To change it, +modify \fBactivation_mode\fP in \fBlvm.conf\fP(5). +. +.HP +.BR \-K | \-\-ignoreactivationskip +.br +Ignore the flag to skip Logical Volumes during activation. +. +.HP +.BR \-k | \-\-setactivationskip +.RB { y | n } +.br Controls whether Logical Volumes are persistently flagged to be skipped during activation. By default, thin snapshot volumes are flagged for activation skip. To activate such volumes, -an extra -.BR \-K / \-\-ignoreactivationskip -option must be used. +an extra \fB\-\-ignoreactivationskip\fP option must be used. The flag is not applied during deactivation. To see whether the flag is attached, use \fBlvs\fP(8) command where the state of the flag is reported within \fBlv_attr\fP bits. -.TP -.BR \-K ", " \-\-ignoreactivationskip -Ignore the flag to skip Logical Volumes during activation. -.TP -.BR \-\-cachepolicy " " \fIpolicy ", " \-\-cachesettings " " \fIkey = \fIvalue +. +.HP +.BR \-\-cachepolicy +.IR policy , +.BR \-\-cachesettings +.IR key \fB= value +.br Only applicable to cached LVs; see also \fBlvmcache(7)\fP. Sets the cache policy and its associated tunable settings. In most use-cases, default values should be adequate. -.TP -.BR \-C ", " \-\-contiguous " {" y | n } +. +.HP +.BR \-C | \-\-contiguous +.RB { y | n } +.br Tries to set or reset the contiguous allocation policy for logical volumes. It's only possible to change a non-contiguous logical volume's allocation policy to contiguous, if all of the allocated physical extents are already contiguous. -.TP +. +.HP .BR \-\-detachprofile +.br Detach any metadata configuration profiles attached to given Logical Volumes. See \fBlvm.conf\fP(5) for more information about metadata profiles. -.TP -.BR \-\-discards " {" ignore | nopassdown | passdown } +. +.HP +.BR \-\-discards +.RB { ignore | nopassdown | passdown } +.br Set this to \fBignore\fP to ignore any discards received by a thin pool Logical Volume. Set to \fBnopassdown\fP to process such discards within the thin pool itself and allow the no-longer-needed extents to be overwritten by new data. Set to \fBpassdown\fP (the default) to process them both within the thin pool itself and to pass them down the underlying device. -.TP -.BR \-\-errorwhenfull " {" y | n } +. +.HP +.BR \-\-errorwhenfull +.RB { y | n } +.br Sets thin pool behavior when data space is exhaused. See .BR lvcreate (8) for information. -.TP -.B \-\-resync -Forces the complete resynchronization of a mirror. In normal -circumstances you should not need this option because synchronization -happens automatically. Data is read from the primary mirror device -and copied to the others, so this can take a considerable amount of -time - and during this time you are without a complete redundant copy -of your data. -.TP -.B \-\-metadataprofile " " \fIProfileName +. +.HP +.BR \-\-ignoremonitoring +.br +Make no attempt to interact with dmeventd unless \fB\-\-monitor\fP +is specified. +Do not use this if dmeventd is already monitoring a device. +. +.HP +.BR \-\-major +.IR major +.br +Sets the major number. This option is supported only on older systems +(kernel version 2.4) and is ignored on modern Linux systems where major +numbers are dynamically assigned. +. +.HP +.BR \-\-minor +.IR minor +.br +Set the minor number. +. +.HP +.BR \-\-metadataprofile +.IR ProfileName +.br Uses and attaches \fIProfileName\fP configuration profile to the logical volume metadata. Whenever the logical volume is processed next time, the profile is automatically applied. If the volume group has another profile attached, the logical volume profile is preferred. See \fBlvm.conf\fP(5) for more information about metadata profiles. -.TP -.B \-\-minor \fIminor -Set the minor number. -.TP -.B \-\-major \fImajor -Sets the major number. This option is supported only on older systems -(kernel version 2.4) and is ignored on modern Linux systems where major -numbers are dynamically assigned. -.TP -.BR \-\-monitor " {" y | n } +. +.HP +.BR \-\-monitor +.RB { y | n } +.br Start or stop monitoring a mirrored or snapshot logical volume with dmeventd, if it is installed. If a device used by a monitored mirror reports an I/O error, the failure is handled according to -\fBmirror_image_fault_policy\fP and \fBmirror_log_fault_policy\fP +\%\fBmirror_image_fault_policy\fP and \fBmirror_log_fault_policy\fP set in \fBlvm.conf\fP(5). -.TP -.BR \-\-poll " {" y | n } +. +.HP +.BR \-\-noudevsync +.br +Disable udev synchronisation. The +process will not wait for notification from udev. +It will continue irrespective of any possible udev processing +in the background. You should only use this if udev is not running +or has rules that ignore the devices LVM2 creates. +. +.HP +.BR \-p | \-\-permission +.RB { r | rw } +.br +Change access permission to read-only or read/write. +. +.HP +.BR \-M | \-\-persistent +.RB { y | n } +.br +Set to \fBy\fP to make the minor number specified persistent. +Change of persistent numbers is not supported for pool volumes. +. +.HP +.BR \-\-poll +.RB { y | n } +.br Without polling a logical volume's backgrounded transformation process will never complete. If there is an incomplete pvmove or lvconvert (for example, on rebooting after a crash), use \fB\-\-poll y\fP to restart the process from its last checkpoint. However, it may not be appropriate to immediately poll a logical volume when it is activated, use \fB\-\-poll n\fP to defer and then \fB\-\-poll y\fP to restart the process. -.TP -.BR \-\- [ raid ] maxrecoveryrate " " \fIRate [ bBsSkKmMgG ] +. +.HP +.BR \-\- [ raid ] maxrecoveryrate +.BR \fIRate [ b | B | s | S | k | K | m | M | g | G ] +.br Sets the maximum recovery rate for a RAID logical volume. \fIRate\fP is specified as an amount per second for each device in the array. If no suffix is given, then KiB/sec/device is assumed. Setting the recovery rate to \fB0\fP means it will be unbounded. -.TP -.BR \-\- [ raid ] minrecoveryrate " " \fIRate [ bBsSkKmMgG ] +. +.HP +.BR \-\- [ raid ] minrecoveryrate +.BR \fIRate [ b | B | s | S | k | K | m | M | g | G ] +.br Sets the minimum recovery rate for a RAID logical volume. \fIRate\fP is specified as an amount per second for each device in the array. If no suffix is given, then KiB/sec/device is assumed. Setting the recovery rate to \fB0\fP means it will be unbounded. -.TP -.BR \-\- [ raid ] syncaction " {" check | repair } +. +.HP +.BR \-\- [ raid ] syncaction +.RB { check | repair } +.br This argument is used to initiate various RAID synchronization operations. The \fBcheck\fP and \fBrepair\fP options provide a way to check the integrity of a RAID logical volume (often referred to as "scrubbing"). @@ -260,16 +338,22 @@ If \fBcheck\fP is used, the discrepancies will be counted but not repaired. If \fBrepair\fP is used, the discrepancies will be corrected as they are encountered. The \fBlvs\fP(8) command can be used to show the number of discrepancies found or repaired. -.TP -.BR \-\- [ raid ] writebehind " " \fIIOCount +. +.HP +.BR \-\- [ raid ] writebehind +.IR IOCount +.br Specify the maximum number of outstanding writes that are allowed to devices in a RAID1 logical volume that are marked as write-mostly. Once this value is exceeded, writes become synchronous (i.e. all writes to the constituent devices must complete before the array signals the write has completed). Setting the value to zero clears the preference and allows the system to choose the value arbitrarily. -.TP -.BR \-\- [ raid ] writemostly " " \fIPhysicalVolume [ : { y | n | t }] +. +.HP +.BR \-\- [ raid ] writemostly +.BR \fIPhysicalVolume [ : { y | n | t }] +.br Mark a device in a RAID1 logical volume as write-mostly. All reads to these drives will be avoided unless absolutely necessary. This keeps the number of I/Os to the drive to a minimum. The default behavior is to @@ -279,8 +363,39 @@ appending a "\fB:n\fP" to the physical volume or to toggle the value by specifyi "\fB:t\fP". The \fB\-\-writemostly\fP argument can be specified more than one time in a single command; making it possible to toggle the write-mostly attributes for all the physical volumes in a logical volume at once. -.TP -.B \-\-sysinit +. +.HP +.BR \-r | \-\-readahead +.RB { \fIReadAheadSectors | auto | none } +.br +Set read ahead sector count of this logical volume. +For volume groups with metadata in lvm1 format, this must +be a value between 2 and 120 sectors. +The default value is "\fBauto\fP" which allows the kernel to choose +a suitable value automatically. +"\fBnone\fP" is equivalent to specifying zero. +. +.HP +.BR \-\-refresh +.br +If the logical volume is active, reload its metadata. +This is not necessary in normal operation, but may be useful +if something has gone wrong or if you're doing clustering +manually without a clustered lock manager. +. +.HP +.BR \-\-resync +.br +Forces the complete resynchronization of a mirror. In normal +circumstances you should not need this option because synchronization +happens automatically. Data is read from the primary mirror device +and copied to the others, so this can take a considerable amount of +time - and during this time you are without a complete redundant copy +of your data. +. +.HP +.BR \-\-sysinit +.br Indicates that \fBlvchange\fP(8) is being invoked from early system initialisation scripts (e.g. rc.sysinit or an initrd), before writeable filesystems are available. As such, @@ -295,51 +410,24 @@ If \fB\-\-sysinit\fP is used in conjunction with \fBlvmetad\fP(8) enabled and running, autoactivation is preferred over manual activation via direct lvchange call. Logical volumes are autoactivated according to -.B auto_activation_volume_list -set in \fBlvm.conf\fP(5). -.TP -.B \-\-noudevsync -Disable udev synchronisation. The -process will not wait for notification from udev. -It will continue irrespective of any possible udev processing -in the background. You should only use this if udev is not running -or has rules that ignore the devices LVM2 creates. -.TP -.B \-\-ignoremonitoring -Make no attempt to interact with dmeventd unless \fB\-\-monitor\fP -is specified. -Do not use this if dmeventd is already monitoring a device. -.TP -.BR \-M ", " \-\-persistent " {" y | n } -Set to \fBy\fP to make the minor number specified persistent. -Change of persistent numbers is not supported for pool volumes. -.TP -.BR \-p ", " \-\-permission " {" r | rw } -Change access permission to read-only or read/write. -.TP -.BR \-r ", " \-\-readahead " {" \fIReadAheadSectors | auto | none } -Set read ahead sector count of this logical volume. -For volume groups with metadata in lvm1 format, this must -be a value between 2 and 120 sectors. -The default value is "\fBauto\fP" which allows the kernel to choose -a suitable value automatically. -"\fBnone\fP" is equivalent to specifying zero. -.TP -.B \-\-refresh -If the logical volume is active, reload its metadata. -This is not necessary in normal operation, but may be useful -if something has gone wrong or if you're doing clustering -manually without a clustered lock manager. -.TP -.BR \-Z ", " \-\-zero " {" y | n } +\fB auto_activation_volume_list\fP set in \fBlvm.conf\fP(5). +. +.HP +.BR \-Z | \-\-zero +.RB { y | n } +.br Set zeroing mode for thin pool. Note: already provisioned blocks from pool in non-zero mode are not cleared in unwritten parts when setting zero to \fBy\fP. +. .SH ENVIRONMENT VARIABLES +. .TP .B LVM_SUPPRESS_LOCKING_FAILURE_MESSAGES Suppress locking failure messages. +. .SH Examples +. Changes the permission on volume lvol1 in volume group vg00 to be read-only: .sp .B lvchange \-pr vg00/lvol1 diff --git a/man/lvcreate.8.in b/man/lvcreate.8.in index ee10392..8a364c4 100644 --- a/man/lvcreate.8.in +++ b/man/lvcreate.8.in @@ -2,40 +2,41 @@ .SH NAME lvcreate \- create a logical volume in an existing volume group .SH SYNOPSIS +.ad l .B lvcreate .RB [ \-a | \-\-activate -.RI [ a | e | l ]{ y | n }] +.RB [ a ][ e | l | s ]{ y | n }] .RB [ \-\-addtag .IR Tag ] .RB [ \-\-alloc -.IR AllocationPolicy ] +.IR Allocation\%Policy ] .RB [ \-A | \-\-autobackup -.RI { y | n }] +.RB { y | n }] .RB [ \-H | \-\-cache ] .RB [ \-\-cachemode -.RI { passthrough | writeback | writethrough }] +.RB { passthrough | writeback | writethrough }] .RB [ \-\-cachepolicy .IR policy ] .RB [ \-\-cachepool -.IR CachePoolLogicalVolume { Name | Path } +.IR CachePoolLogicalVolume ] .RB [ \-\-cachesettings -.IR key=value ] +.IR key \fB= value ] .RB [ \-c | \-\-chunksize -.IR ChunkSize [ bBsSkKmMgG ]] +.IR ChunkSize ] .RB [ \-\-commandprofile .IR ProfileName ] .RB [ \-C | \-\-contiguous -.RI { y | n }] +.RB { y | n }] .RB [ \-d | \-\-debug ] .RB [ \-\-discards -.RI { ignore | nopassdown | passdown }] +.RB \%{ ignore | nopassdown | passdown }] .RB [ \-\-errorwhenfull -.RI { y | n }] +.RB { y | n }] .RB [{ \-l | \-\-extents -.IR LogicalExtentsNumber [ % { FREE | PVS | VG }] -| +.BR \fILogicalExtents\%Number [ % { FREE | PVS | VG }] +.RB | .BR \-L | \-\-size -.IR LogicalVolumeSize [ bBsSkKmMgGtTpPeE ]} +.BR \fILogicalVolumeSize } .RB [ \-i | \-\-stripes .IR Stripes .RB [ \-I | \-\-stripesize @@ -48,86 +49,79 @@ lvcreate \- create a logical volume in an existing volume group .RB [ \-j | \-\-major .IR major ]] .RB [ \-\-metadataprofile -.IR ProfileName ] +.IR Profile\%Name ] .RB [ \-m | \-\-mirrors .IR Mirrors -.RB [{ \-\-corelog -| -.B \-\-mirrorlog -.RI { disk | core | mirrored }}] +.RB [ \-\-corelog | \-\-mirrorlog +.RB { disk | core | mirrored }] .RB [ \-\-nosync ] .RB [ \-R | \-\-regionsize -.IR MirrorLogRegionSize [ bBsSkKmMgG ]]] +.BR \fIMirrorLogRegionSize ]] .RB [ \-\-monitor -.RI { y | n }] +.RB { y | n }] .RB [ \-n | \-\-name -.IR LogicalVolume { Name | Path }] +.IR Logical\%Volume ] .RB [ \-\-noudevsync ] .RB [ \-p | \-\-permission -.RI { r | rw }] +.RB { r | rw }] .RB [ \-M | \-\-persistent -.RI { y | n }] +.RB { y | n }] .\" .RB [ \-\-pooldatasize -.\" .IR DataVolumeSize [ bBsSkKmMgGtTpPeE ]] -.RB [ \-\-poolmetadatasize -.IR MetadataVolumeSize [ bBsSkKmMgG ]] +.\" .I DataVolumeSize +.RB \%[ \-\-poolmetadatasize +.IR MetadataVolumeSize ] .RB [ \-\-poolmetadataspare -.RI { y | n }] +.RB { y | n }] .RB [ \-\- [ raid ] maxrecoveryrate .IR Rate ] .RB [ \-\- [ raid ] minrecoveryrate .IR Rate ] .RB [ \-r | \-\-readahead -.RI { ReadAheadSectors | auto | none }] +.RB { \fIReadAheadSectors | auto | none }] .RB [ \-k | \-\-setactivationskip -.RI { y | n }] -.RB [ \-s | \-\-snapshot +.RB { y | n }] +.RB [ \-s | \-\-snapshot ] .RB [ \-V | \-\-virtualsize -.IR VirtualSize [ bBsSkKmMgGtTpPeE ]] +.IR VirtualSize ] .RB [ \-t | \-\-test ] .RB [ \-T | \-\-thin ] .RB [ \-\-thinpool -.IR ThinPoolLogicalVolume { Name | Path }] +.IR ThinPoolLogicalVolume ] .RB [ \-\-type .IR SegmentType ] .RB [ \-v | \-\-verbose ] .RB [ \-W | \-\-wipesignatures -.RI { y | n }] +.RB { y | n }] .RB [ \-Z | \-\-zero -.RI { y | n }] -.RI [ VolumeGroup { Name | Path } -.RI [/\{ ExternalOrigin -| -.I Origin -| +.RB { y | n }] +.RI [ VolumeGroup [ \fB/ { ExternalOrigin |\: Origin | .IR Pool } LogicalVolumeName ] -.RI [ PhysicalVolumePath [ :PE [ \-PE ]]...] - +.RI [ PhysicalVolumePath [ \fB: \fIPE \fR[ \fB\- PE ]]...] +.LP .B lvcreate .RB [ \-l | \-\-extents -.IR LogicalExtentsNumber [ % { FREE | ORIGIN | PVS | VG }] +.BR \fILogicalExtentsNumber [ % { FREE | ORIGIN | PVS | VG }] | .BR \-L | \-\-size .\" | \-\-pooldatasize -.IR LogicalVolumeSize [ bBsSkKmMgGtTpPeE ]] +.IR LogicalVolumeSize ] .RB [ \-c | \-\-chunksize -.IR ChunkSize [ bBsSkKmMgG ]] +.IR ChunkSize ] .RB [ \-\-commandprofile -.IR Profilename ] +.IR Profile\%Name ] .RB [ \-\-noudevsync ] .RB [ \-\-ignoremonitoring ] -.RB [ \-\-metadataProfile -.IR ProfileName ] -.RB [ \-\-monitor -.RI { y | n }] +.RB [ \-\-metadataprofile +.IR Profile\%Name ] +.RB \%[ \-\-monitor +.RB { y | n }] .RB [ \-n | \-\-name -.IR SnapshotLogicalVolume { Name | Path }] +.IR SnapshotLogicalVolume ] .BR \-s | \-\-snapshot | \-H | \-\-cache -.RI {[ VolumeGroup { Name | Path }/] OriginalLogicalVolumeName +.RI \%{[ VolumeGroup \fB/ ] OriginalLogicalVolume .BR \-V | \-\-virtualsize -.IR VirtualSize [ bBsSkKmMgGtTpPeE ]} -.br - +.IR VirtualSize } +.ad b .SH DESCRIPTION lvcreate creates a new logical volume in a volume group (see .BR vgcreate "(8), " vgchange (8)) @@ -144,12 +138,17 @@ extents will be restricted to these volumes. .br The second form supports the creation of snapshot logical volumes which keep the contents of the original logical volume for backup purposes. +. .SH OPTIONS +. See .BR lvm (8) for common options. -.TP -.IR \fB\-a ", " \fB\-\-activate " {" y | ay | n | ey | en | ly | ln } +. +.HP +.BR \-a | \-\-activate +.RB [ a ][ l | e | s ]{ y | n } +.br Controls the availability of the Logical Volumes for immediate use after the command finishes running. By default, new Logical Volumes are activated (\fB\-a\fIy\fR). @@ -171,8 +170,10 @@ be overridden. If the clustered locking is enabled, \fB\-a\fIey\fR will activate exclusively on one node and .IR \fB\-a { a | l } y will activate only on the local node. -.TP -.IR \fB\-H ", " \fB\-\-cache +. +.HP +.BR \-H | \-\-cache +.br Creates cache or cache pool logical volume or both. Specifying the optional argument \fB\-\-size\fP will cause the creation of the cache logical volume. @@ -187,33 +188,49 @@ the \fB\-\-size\fP is used to specify size of cache pool volume. See \fBlvmcache\fP(7) for more info about caching support. Note that the cache segment type requires a dm-cache kernel module version 1.3.0 or greater. -.TP -.IR \fB\-\-cachemode " {" passthrough | writeback | writethrough } +. +.HP +.BR \-\-cachemode +.RB { passthrough | writeback | writethrough } +.br Specifying a cache mode determines when the writes to a cache LV -are considered complete. When \fIwriteback\fP is specified, a write is +are considered complete. When \fBwriteback\fP is specified, a write is considered complete as soon as it is stored in the cache pool LV. -If \fIwritethough\fP is specified, a write is considered complete only +If \fBwritethough\fP is specified, a write is considered complete only when it has been stored in the cache pool LV and on the origin LV. -While \fIwritethrough\fP may be slower for writes, it is more +While \fBwritethrough\fP may be slower for writes, it is more resilient if something should happen to a device associated with the cache pool LV. -.TP -.B \-\-cachepolicy \fIpolicy +. +.HP +.BR \-\-cachepolicy +.IR policy +.br Only applicable to cached LVs; see also \fBlvmcache(7)\fP. Sets -the cache policy. \fImq\fP is the basic policy name. \fIsmq\fP is more advanced +the cache policy. \fBmq\fP is the basic policy name. \fBsmq\fP is more advanced version available in newer kernels. -.TP -.IR \fB\-\-cachepool " " CachePoolLogicalVolume { Name | Path } +. +.HP +.BR \-\-cachepool +.IR CachePoolLogicalVolume { Name | Path } +.br Specifies the name of cache pool volume name. The other way to specify pool name is to append name to Volume group name argument. -.TP -.BR \-\-cachesettings " " \fIkey=value +. +.HP +.BR \-\-cachesettings +.IB key = value +.br Only applicable to cached LVs; see also \fBlvmcache(7)\fP. Sets the cache tunable settings. In most use-cases, default values should be adequate. -Special string value \fIdefault\fP switches setting back to its default kernel value +Special string value \fBdefault\fP switches setting back to its default kernel value and removes it from the list of settings stored in lvm2 metadata. -.TP -.BR \-c ", " \-\-chunksize " " \fIChunkSize [ \fIbBsSkKmMgG ] +. +.HP +.BR \-c | \-\-chunksize +.IR ChunkSize \c +.RB [ b | B | s | S | k | K | m | M | g | G ] +.br Gives the size of chunk for snapshot, cache pool and thin pool logical volumes. Default unit is in kilobytes. .br @@ -230,79 +247,108 @@ fit the pool metadata size within 128MiB, if the pool metadata size is not specified. See .BR lvm.conf (5) -setting -.IR allocation / thin_pool_chunk_size_policy +setting \fBallocation/thin_pool_chunk_size_policy\fP to select different calculation policy. Thin pool target version <1.4 requires this value to be a power of 2. For target version <1.5 discard is not supported for non power of 2 values. -.TP -.BR \-C ", " \-\-contiguous " {" \fIy | \fIn } +. +.HP +.BR \-C | \-\-contiguous +.RB { y | n } +.br Sets or resets the contiguous allocation policy for logical volumes. Default is no contiguous allocation based on a next free principle. -.TP +. +.HP .BR \-\-corelog -This is shortcut for option \fB\-\-mirrorlog\fP \fIcore\fP. -.TP -.BR \-\-discards " {" \fIignore | \fInopassdown | \fIpassdown } +.br +This is shortcut for option \fB\-\-mirrorlog core\fP. +. +.HP +.BR \-\-discards +.RB { ignore | nopassdown | passdown } +.br Sets discards behavior for thin pool. -Default is \fIpassdown\fP. -.TP -.BR \-\-errorwhenfull " {" \fIy |\fIn } +Default is \fBpassdown\fP. +. +.HP +.BR \-\-errorwhenfull +.RB { y | n } +.br Configures thin pool behaviour when data space is exhausted. -Default is \fIn\fPo. +Default is \fBn\fPo. Device will queue I/O operations until target timeout (see dm-thin-pool kernel module option \fIno_space_timeout\fP) expires. Thus configured system has a time to i.e. extend the size of thin pool data device. -When set to \fIy\fPes, the I/O operation is immeditelly errored. -.TP -.BR \-K ", " \-\-ignoreactivationskip +When set to \fBy\fPes, the I/O operation is immeditelly errored. +. +.HP +.BR \-K | \-\-ignoreactivationskip +.br Ignore the flag to skip Logical Volumes during activation. Use \fB\-\-setactivationskip\fP option to set or reset activation skipping flag persistently for logical volume. -.TP -.B \-\-ignoremonitoring +. +.HP +.BR \-\-ignoremonitoring +.br Make no attempt to interact with dmeventd unless \fB\-\-monitor\fP is specified. -.TP -.IR \fB\-l ", " \fB\-\-extents " " LogicalExtentsNumber [ % { VG | PVS | FREE | ORIGIN }] +. +.HP +.BR -l | \-\-extents +.IR LogicalExtentsNumber \c +.RB [ % { VG | PVS | FREE | ORIGIN }] +.br Gives the number of logical extents to allocate for the new logical volume. The total number of physical extents allocated will be greater than this, for example, if the volume is mirrored. The number can also be expressed as a percentage of the total space -in the Volume Group with the suffix \fI%VG\fR, as a percentage of the -remaining free space in the Volume Group with the suffix \fI%FREE\fR, as a +in the Volume Group with the suffix \fB%VG\fP, as a percentage of the +remaining free space in the Volume Group with the suffix \fB%FREE\fP, as a percentage of the remaining free space for the specified -PhysicalVolume(s) with the suffix \fI%PVS\fR, or (for a snapshot) as a +PhysicalVolume(s) with the suffix \fB%PVS\fP, or (for a snapshot) as a percentage of the total space in the Origin Logical Volume with the -suffix \fI%ORIGIN\fR (i.e. \fI100%ORIGIN\fR provides space for the whole origin). +suffix \fB%ORIGIN\fP (i.e. \fB100%ORIGIN\fP provides space for the whole origin). When expressed as a percentage, the number is treated as an approximate upper limit for the total number of physical extents to be allocated (including extents used by any mirrors, for example). -.TP -.BR \-j ", " \-\-major " " \fImajor +. +.HP +.BR \-j | \-\-major +.IR major +.br Sets the major number. Major numbers are not supported with pool volumes. This option is supported only on older systems (kernel version 2.4) and is ignored on modern Linux systems where major numbers are dynamically assigned. -.TP -.BR \-\-metadataprofile " " \fIProfileName +. +.HP +.BR \-\-metadataprofile +.IR ProfileName +.br Uses and attaches the ProfileName configuration profile to the logical volume metadata. Whenever the logical volume is processed next time, the profile is automatically applied. If the volume group has another profile attached, the logical volume profile is preferred. See \fBlvm.conf\fP(5) for more information about \fBmetadata profiles\fP. -.TP -.B \-\-minor \fIminor +. +.HP +.BR \-\-minor +.IR minor +.br Sets the minor number. Minor numbers are not supported with pool volumes. -.TP -.BR \-m ", " \-\-mirrors " " \fIMirrors -Creates a mirrored logical volume with \fIMirrors\fP copies. -For example, specifying -.BI \-m\ 1 +. +.HP +.BR \-m | \-\-mirrors +.IR mirrors +.br +Creates a mirrored logical volume with \fImirrors\fP copies. +For example, specifying \fB\-m 1\fP would result in a mirror with two-sides; that is, a linear volume plus one copy. @@ -318,131 +364,173 @@ The default is "\fIraid1\fP". See the \fB\-\-type\fP option for more information if you would like to use the legacy "\fImirror\fP" segment type. See .BR lvm.conf (5) -settings -.IR global / mirror_segtype_default -and -.IR global / raid10_segtype_default +settings \fB global/mirror_segtype_default\fP +and \fBglobal/raid10_segtype_default\fP to configure default mirror segment type. The options \fB\-\-mirrorlog\fP and \fB\-\-corelog\fP apply to the legacy "\fImirror\fP" segment type only. -.TP -.BR \-\-mirrorlog " {" \fIdisk | \fIcore | \fImirrored } +. +.HP +.BR \-\-mirrorlog +.RB { disk | core | mirrored } +.br Specifies the type of log to be used for logical volumes utilizing the legacy "\fImirror\fP" segment type. .br -The default is \fIdisk\fP, which is persistent and requires +The default is \fBdisk\fP, which is persistent and requires a small amount of storage space, usually on a separate device from the data being mirrored. .br -Using \fIcore\fP means the mirror is regenerated by copying the data +Using \fBcore\fP means the mirror is regenerated by copying the data from the first device each time the logical volume is activated, like after every reboot. .br -Using \fImirrored\fP will create a persistent log that is itself mirrored. -.TP -.BR \-\-monitor " {" \fIy | \fIn } +Using \fBmirrored\fP will create a persistent log that is itself mirrored. +. +.HP +.BR \-\-monitor +.RB { y | n } +.br Starts or avoids monitoring a mirrored, snapshot or thin pool logical volume with dmeventd, if it is installed. If a device used by a monitored mirror reports an I/O error, the failure is handled according to -.IR activation / mirror_image_fault_policy -and -.IR activation / mirror_log_fault_policy +\fBactivation/mirror_image_fault_policy\fP +and \fBactivation/mirror_log_fault_policy\fP set in \fBlvm.conf\fP(5). -.TP -.IR \fB\-n ", " \fB\-\-name " " LogicalVolume { Name | Path } +. +.HP +.BR \-n | \-\-name +.IR LogicalVolume { Name | Path } +.br Sets the name for the new logical volume. .br Without this option a default name of "lvol#" will be generated where # is the LVM internal number of the logical volume. -.TP -.B \-\-nosync +. +.HP +.BR \-\-nosync +.br Causes the creation of the mirror to skip the initial resynchronization. -.TP -.B \-\-noudevsync +. +.HP +.BR \-\-noudevsync +.br Disables udev synchronisation. The process will not wait for notification from udev. It will continue irrespective of any possible udev processing in the background. You should only use this if udev is not running or has rules that ignore the devices LVM2 creates. -.TP -.BR \-p ", " \-\-permission " {" \fIr | \fIrw } -Sets access permissions to read only (\fIr\fP) or read and write (\fIrw\fP). +. +.HP +.BR \-p | \-\-permission +.RB { r | rw } +.br +Sets access permissions to read only (\fBr\fP) or read and write (\fBrw\fP). .br Default is read and write. -.TP -.BR \-M ", " \-\-persistent " {" \fIy | \fIn } -Set to \fIy\fP to make the minor number specified persistent. +. +.HP +.BR \-M | \-\-persistent +.RB { y | n } +.br +Set to \fBy\fP to make the minor number specified persistent. Pool volumes cannot have persistent major and minor numbers. -Defaults to \fIy\fPes only when major or minor number is specified. -Otherwise it is \fIn\fPo. -.TP +Defaults to \fBy\fPes only when major or minor number is specified. +Otherwise it is \fBn\fPo. +.\" .HP .\" .IR \fB\-\-pooldatasize " " PoolDataVolumeSize [ bBsSkKmMgGtTpPeE ] .\" Sets the size of pool's data logical volume. .\" For thin pools you may also specify the size .\" with the option \fB\-\-size\fP. -.\" .TP -.IR \fB\-\-poolmetadatasize " " MetadataVolumeSize [ bBsSkKmMgG ] +.\" +. +.HP +.BR \-\-poolmetadatasize +.BR \fIMetadataVolumeSize [ b | B | s | S | k | K | m | M | g | G ] +.br Sets the size of pool's metadata logical volume. Supported values are in range between 2MiB and 16GiB for thin pool, and upto 16GiB for cache pool. The minimum value is computed from pool's data size. Default value for thin pool is (Pool_LV_size / Pool_LV_chunk_size * 64b). Default unit is megabytes. -.TP -.IR \fB\-\-poolmetadataspare " {" y | n } +. +.HP +.BR \-\-poolmetadataspare +.RB { y | n } +.br Controls creation and maintanence of pool metadata spare logical volume that will be used for automated pool recovery. Only one such volume is maintained within a volume group with the size of the biggest pool metadata volume. -Default is \fIy\fPes. -.TP -.BR \-\- [ raid ] maxrecoveryrate " " \fIRate [ \fIbBsSkKmMgG ] +Default is \fBy\fPes. +. +.HP +.BR \-\- [ raid ] maxrecoveryrate +.BR \fIRate [ b | B | s | S | k | K | m | M | g | G ] +.br Sets the maximum recovery rate for a RAID logical volume. \fIRate\fP is specified as an amount per second for each device in the array. If no suffix is given, then KiB/sec/device is assumed. Setting the recovery rate to 0 means it will be unbounded. -.TP -.BR \-\- [ raid ] minrecoveryrate " " \fIRate [ \fIbBsSkKmMgG ] +. +.HP +.BR \-\- [ raid ] minrecoveryrate +.BR \fIRate [ b | B | s | S | k | K | m | M | g | G ] +.br Sets the minimum recovery rate for a RAID logical volume. \fIRate\fP is specified as an amount per second for each device in the array. If no suffix is given, then KiB/sec/device is assumed. Setting the recovery rate to 0 means it will be unbounded. -.TP -.IR \fB\-r ", " \fB\-\-readahead " {" ReadAheadSectors | auto | none } +. +.HP +.BR \-r | \-\-readahead +.RB { \fIReadAheadSectors | auto | none } +.br Sets read ahead sector count of this logical volume. For volume groups with metadata in lvm1 format, this must be a value between 2 and 120. -The default value is \fIauto\fP which allows the kernel to choose +The default value is \fBauto\fP which allows the kernel to choose a suitable value automatically. -\fINone\fP is equivalent to specifying zero. -.TP -.BR \-R ", " \-\-regionsize " " \fIMirrorLogRegionSize [ \fIbBsSkKmMgG ] +\fBnone\fP is equivalent to specifying zero. +. +.HP +.BR \-R | \-\-regionsize +.BR \fIMirrorLogRegionSize [ b | B | s | S | k | K | m | M | g | G ] +.br A mirror is divided into regions of this size (in MiB), and the mirror log uses this granularity to track which regions are in sync. -.TP -.IR \fB\-k ", " \fB\-\-setactivationskip " {" y | n } +. +.HP +.BR \-k | \-\-setactivationskip +.RB { y | n } +.br Controls whether Logical Volumes are persistently flagged to be skipped during activation. By default, thin snapshot volumes are flagged for activation skip. See .BR lvm.conf (5) -.IR activation / auto_set_activation_skip +\fBactivation/auto_set_activation_skip\fP how to change its default behaviour. -To activate such volumes, an extra -.BR \-K | \-\-ignoreactivationskip +To activate such volumes, an extra \fB\-\-ignoreactivationskip\fP option must be used. The flag is not applied during deactivation. Use -.B lvchange \-\-setactivationskip {y|n} +\fBlvchange \-\-setactivationskip\fP command to change the skip flag for existing volumes. To see whether the flag is attached, use \fBlvs\fP command where the state of the flag is reported within \fBlv_attr\fP bits. -.TP -.IR \fB\-L ", " \fB\-\-size " " LogicalVolumeSize [ bBsSkKmMgGtTpPeE ] +. +.HP +.BR \-L | \-\-size +.BR \fILogicalVolumeSize [ b | B | s | S | k | K | m | M | \c +.BR g | G | t | T | p | P | e | E ] +.\" man2html cannot handle all those changes in 1 line +.br Gives the size to allocate for the new logical volume. -A size suffix of \fIB\fR for bytes, \fIS\fR for sectors as 512 bytes, -\fIK\fR for kilobytes, \fIM\fR for megabytes, -\fIG\fR for gigabytes, \fIT\fR for terabytes, \fIP\fR for petabytes -or \fIE\fR for exabytes is optional. +A size suffix of \fBB\fP for bytes, \fBS\fP for sectors as 512 bytes, +\fBK\fP for kilobytes, \fBM\fP for megabytes, +\fBG\fP for gigabytes, \fBT\fP for terabytes, \fBP\fP for petabytes +or \fBE\fP for exabytes is optional. .br Default unit is megabytes. .TP @@ -589,7 +677,7 @@ If blkid wiping is used setting in .BR lvm.conf (5)) and LVM2 is compiled with blkid wiping support, then \fBblkid\fP(8) library is used -to detect the signatures (use \fBblkid -k\fP command to list the signatures that are recognized). +to detect the signatures (use \fBblkid \-k\fP command to list the signatures that are recognized). Otherwise, native LVM2 code is used to detect signatures (MD RAID, swap and LUKS signatures are detected only in this case). .br @@ -705,7 +793,7 @@ volume (i.e. the origin LV), creating a cache LV. .\" Create a 1G cached LV "lvol1" with 10M cache pool "vg00/pool". .\" .sp -.\" .B lvcreate \-\-cache \-L 1G \-n lv --pooldatasize 10M vg00/pool +.\" .B lvcreate \-\-cache \-L 1G \-n lv \-\-pooldatasize 10M vg00/pool .SH SEE ALSO .BR lvm (8), diff --git a/man/lvm.8.in b/man/lvm.8.in index daab791..8ff31ae 100644 --- a/man/lvm.8.in +++ b/man/lvm.8.in @@ -1,29 +1,35 @@ .TH LVM 8 "LVM TOOLS #VERSION#" "Sistina Software UK" \" -*- nroff -*- +. .SH NAME +. lvm \(em LVM2 tools +. .SH SYNOPSIS +. .B lvm -[command | file] +.RI [ command | file ] +. .SH DESCRIPTION +. lvm provides the command-line tools for LVM2. A separate manual page describes each command in detail. -.LP +.P If \fBlvm\fP is invoked with no arguments it presents a readline prompt (assuming it was compiled with readline support). LVM commands may be entered interactively at this prompt with readline facilities including history and command name and option completion. Refer to \fBreadline\fP(3) for details. -.LP +.P If \fBlvm\fP is invoked with argv[0] set to the name of a specific LVM command (for example by using a hard or soft link) it acts as that command. -.LP +.P On invocation, \fBlvm\fP requires that only the standard file descriptors stdin, stdout and stderr are available. If others are found, they get closed and messages are issued warning about the leak. This warning can be suppressed by setting the environment variable .B LVM_SUPPRESS_FD_WARNINGS\fP. -.LP +.P Where commands take VG or LV names as arguments, the full path name is optional. An LV called "lvol0" in a VG called "vg0" can be specified as "vg0/lvol0". Where a list of VGs is required but is left empty, @@ -31,223 +37,319 @@ a list of all VGs will be substituted. Where a list of LVs is required but a VG is given, a list of all the LVs in that VG will be substituted. So \fBlvdisplay vg0\fP will display all the LVs in "vg0". Tags can also be used - see \fB\-\-addtag\fP below. -.LP +.P One advantage of using the built-in shell is that configuration information gets cached internally between commands. -.LP +.P A file containing a simple script with one command per line can also be given on the command line. The script can also be executed directly if the first line is #! followed by the absolute path of \fBlvm\fP. +. .SH BUILT-IN COMMANDS -The following commands are built into lvm without links normally -being created in the filesystem for them. +. +The following commands are built into lvm without links +normally being created in the filesystem for them. +.sp .PD 0 -.HP -\fBconfig\fP \(em The same as \fBlvmconfig\fP(8) below. -.HP -\fBdevtypes\fP \(em Display the recognised built-in block device types. -.HP -\fBdumpconfig\fP \(em The same as \fBlvmconfig\fP(8) below. -.HP -\fBformats\fP \(em Display recognised metadata formats. -.HP -\fBhelp\fP \(em Display the help text. -.HP -\fBlvpoll\fP \(em Internal command used by lvmpolld to complete some Logical Volume operations. -.HP -\fBpvdata\fP \(em Not implemented in LVM2. -.HP -\fBsegtypes\fP \(em Display recognised Logical Volume segment types. -.HP -\fBsystemid\fP \(em Display the system ID, if any, currently set on this host. -.HP -\fBtags\fP \(em Display any tags defined on this host. -.HP -\fBversion\fP \(em Display version information. +.TP 14 +.B config +The same as \fBlvmconfig\fP(8) below. +.TP +.B devtypes +Display the recognised built-in block device types. +.TP +.B dumpconfig +The same as \fBlvmconfig\fP(8) below. +.TP +.B formats +Display recognised metadata formats. +.TP +.B help +Display the help text. +.TP +.B lvpoll +Complete lvmpolld operations (Internal command). +.TP +.B pvdata +Not implemented in LVM2. +.TP +.B segtypes +Display recognised Logical Volume segment types. +.TP +.B systemid +Display any system ID currently set on this host. +.TP +.B tags +Display any tags defined on this host. +.TP +.B version +Display version information. .PD +. .SH COMMANDS +. The following commands implement the core LVM functionality. +.sp .PD 0 -.HP -\fBpvchange\fP \(em Change attributes of a Physical Volume. -.HP -\fBpvck\fP \(em Check Physical Volume metadata. -.HP -\fBpvcreate\fP \(em Initialize a disk or partition for use by LVM. -.HP -\fBpvdisplay\fP \(em Display attributes of a Physical Volume. -.HP -\fBpvmove\fP \(em Move Physical Extents. -.HP -\fBpvremove\fP \(em Remove a Physical Volume. -.HP -\fBpvresize\fP \(em Resize a disk or partition in use by LVM2. -.HP -\fBpvs\fP \(em Report information about Physical Volumes. -.HP -\fBpvscan\fP \(em Scan all disks for Physical Volumes. -.HP -\fBvgcfgbackup\fP \(em Backup Volume Group descriptor area. -.HP -\fBvgcfgrestore\fP \(em Restore Volume Group descriptor area. -.HP -\fBvgchange\fP \(em Change attributes of a Volume Group. -.HP -\fBvgck\fP \(em Check Volume Group metadata. -.HP -\fBvgconvert\fP \(em Convert Volume Group metadata format. -.HP -\fBvgcreate\fP \(em Create a Volume Group. -.HP -\fBvgdisplay\fP \(em Display attributes of Volume Groups. -.HP -\fBvgexport\fP \(em Make volume Groups unknown to the system. -.HP -\fBvgextend\fP \(em Add Physical Volumes to a Volume Group. -.HP -\fBvgimport\fP \(em Make exported Volume Groups known to the system. -.HP -\fBvgimportclone\fP \(em Import and rename duplicated Volume Group (e.g. a hardware snapshot). -.HP -\fBvgmerge\fP \(em Merge two Volume Groups. -.HP -\fBvgmknodes\fP \(em Recreate Volume Group directory and Logical Volume special files -.HP -\fBvgreduce\fP \(em Reduce a Volume Group by removing one or more -Physical Volumes. -.HP -\fBvgremove\fP \(em Remove a Volume Group. -.HP -\fBvgrename\fP \(em Rename a Volume Group. -.HP -\fBvgs\fP \(em Report information about Volume Groups. -.HP -\fBvgscan\fP \(em Scan all disks for Volume Groups and rebuild caches. -.HP -\fBvgsplit\fP \(em Split a Volume Group into two, moving any logical +.TP 14 +.B pvchange +Change attributes of a Physical Volume. +.TP +.B pvck +Check Physical Volume metadata. +.TP +.B pvcreate +Initialize a disk or partition for use by LVM. +.TP +.B pvdisplay +Display attributes of a Physical Volume. +.TP +.B pvmove +Move Physical Extents. +.TP +.B pvremove +Remove a Physical Volume. +.TP +.B pvresize +Resize a disk or partition in use by LVM2. +.TP +.B pvs +Report information about Physical Volumes. +.TP +.B pvscan +Scan all disks for Physical Volumes. +.TP +.B vgcfgbackup +Backup Volume Group descriptor area. +.TP +.B vgcfgrestore +Restore Volume Group descriptor area. +.TP +.B vgchange +Change attributes of a Volume Group. +.TP +.B vgck +Check Volume Group metadata. +.TP +.B vgconvert +Convert Volume Group metadata format. +.TP +.B vgcreate +Create a Volume Group. +.TP +.B vgdisplay +Display attributes of Volume Groups. +.TP +.B vgexport +Make volume Groups unknown to the system. +.TP +.B vgextend +Add Physical Volumes to a Volume Group. +.TP +.B vgimport +Make exported Volume Groups known to the system. +.TP +.B vgimportclone +Import and rename duplicated Volume Group (e.g. a hardware snapshot). +.TP +.B vgmerge +Merge two Volume Groups. +.TP +.B vgmknodes +Recreate Volume Group directory and Logical Volume special files +.TP +.B vgreduce +Reduce a Volume Group by removing one or more Physical Volumes. +.TP +.B vgremove +Remove a Volume Group. +.TP +.B vgrename +Rename a Volume Group. +.TP +.B vgs +Report information about Volume Groups. +.TP +.B vgscan +Scan all disks for Volume Groups and rebuild caches. +.TP +.B vgsplit +Split a Volume Group into two, moving any logical volumes from one Volume Group to another by moving entire Physical Volumes. -.HP -\fBlvchange\fP \(em Change attributes of a Logical Volume. -.HP -\fBlvconvert\fP \(em Convert a Logical Volume from linear to mirror or snapshot. -.HP -\fBlvcreate\fP \(em Create a Logical Volume in an existing Volume Group. -.HP -\fBlvdisplay\fP \(em Display attributes of a Logical Volume. -.HP -\fBlvextend\fP \(em Extend the size of a Logical Volume. -.HP -\fBlvmchange\fP \(em Change attributes of the Logical Volume Manager. -.HP -\fBlvmconfig\fP \(em Display the configuration information after +.TP +.B lvchange +Change attributes of a Logical Volume. +.TP +.B lvconvert +Convert a Logical Volume from linear to mirror or snapshot. +.TP +.B lvcreate +Create a Logical Volume in an existing Volume Group. +.TP +.B lvdisplay +Display attributes of a Logical Volume. +.TP +.B lvextend +Extend the size of a Logical Volume. +.TP +.B lvmchange +Change attributes of the Logical Volume Manager. +.TP +.B lvmconfig +Display the configuration information after loading \fBlvm.conf\fP(5) and any other configuration files. -.HP -\fBlvmdiskscan\fP \(em Scan for all devices visible to LVM2. -.HP -\fBlvmdump\fP \(em Create lvm2 information dumps for diagnostic purposes. -.HP -\fBlvreduce\fP \(em Reduce the size of a Logical Volume. -.HP -\fBlvremove\fP \(em Remove a Logical Volume. -.HP -\fBlvrename\fP \(em Rename a Logical Volume. -.HP -\fBlvresize\fP \(em Resize a Logical Volume. -.HP -\fBlvs\fP \(em Report information about Logical Volumes. -.HP -\fBlvscan\fP \(em Scan (all disks) for Logical Volumes. +.TP +.B lvmdiskscan +Scan for all devices visible to LVM2. +.TP +.B lvmdump +Create lvm2 information dumps for diagnostic purposes. +.TP +.B lvreduce +Reduce the size of a Logical Volume. +.TP +.B lvremove +Remove a Logical Volume. +.TP +.B lvrename +Rename a Logical Volume. +.TP +.B lvresize +Resize a Logical Volume. +.TP +.B lvs +Report information about Logical Volumes. +.TP +.B lvscan +Scan (all disks) for Logical Volumes. .PD -.HP -The following commands are not implemented in LVM2 but might be in the future: lvmsadc, lvmsar, pvdata. +.P +The following commands are not implemented in LVM2 but might be +in the future: +.BR lvmsadc ", " lvmsar ", " pvdata . +. .SH OPTIONS +. The following options are available for many of the commands. They are implemented generically and documented here rather than repeated on individual manual pages. - +.P Additional hyphens within option names are ignored. For example, -\-\-readonly and \-\-read\-only are both accepted. - -.TP -.BR \-h ", " \-? ", " \-\-help +\fB\-\-readonly\fP and \fB\-\-read\-only\fP are both accepted. +. +.HP +.BR \-h | \-? | \-\-help +.br Display the help text. -.TP -.B \-\-version +. +.HP +.BR \-\-version +.br Display version information. -.TP -.BR \-v ", " \-\-verbose +. +.HP +.BR \-v | \-\-verbose +.br Set verbose level. Repeat from 1 to 3 times to increase the detail of messages sent to stdout and stderr. Overrides config file setting. -.TP -.BR \-d ", " \-\-debug +. +.HP +.BR \-d | \-\-debug +.br Set debug level. Repeat from 1 to 6 times to increase the detail of messages sent to the log file and/or syslog (if configured). Overrides config file setting. -.TP -.BR \-q ", " \-\-quiet +. +.HP +.BR \-q | \-\-quiet +.br Suppress output and log messages. Overrides \fB\-d\fP and \fB\-v\fP. Repeat once to also suppress any prompts with answer 'no'. -.TP +. +.HP .BR \-\-yes +.br Don't prompt for confirmation interactively but instead always assume the answer is 'yes'. Take great care if you use this! -.TP -.BR \-t ", " \-\-test +. +.HP +.BR \-t | \-\-test +.br Run in test mode. Commands will not update metadata. This is implemented by disabling all metadata writing but nevertheless returning success to the calling function. This may lead to unusual error messages in multi-stage operations if a tool relies on reading back metadata it believes has changed but hasn't. -.TP -.BR \-\-driverloaded " {" \fIy | \fIn } +. +.HP +.BR \-\-driverloaded +.RB { y | n } +.br Whether or not the device-mapper kernel driver is loaded. -If you set this to \fIn\fP, no attempt will be made to contact the driver. -.TP -.BR \-A ", " \-\-autobackup " {" \fIy | \fIn } +If you set this to \fBn\fP, no attempt will be made to contact the driver. +. +.HP +.BR \-A | \-\-autobackup +.RB { y | n } +.br Whether or not to metadata should be backed up automatically after a change. You are strongly advised not to disable this! See \fBvgcfgbackup\fP(8). -.TP -.BR \-P ", " \-\-partial +. +.HP +.BR \-P | \-\-partial +.br When set, the tools will do their best to provide access to Volume Groups that are only partially available (one or more Physical Volumes belonging to the Volume Group are missing from the system). Where part of a logical -volume is missing, \fB/dev/ioerror\fP will be substituted, and you could use +volume is missing, \fI\%/dev/ioerror\fP will be substituted, and you could use \fBdmsetup\fP(8) to set this up to return I/O errors when accessed, or create it as a large block device of nulls. Metadata may not be changed with this option. To insert a replacement Physical Volume of the same or large size use \fBpvcreate \-u\fP to set the uuid to match the original followed by \fBvgcfgrestore\fP(8). -.TP -.BR \-S ", " \-\-select " " \fISelection -For reporting commands, display only rows that match selection criteria. +. +.HP +.BR \-S | \-\-select +.IR Selection +.br +For reporting commands, display only rows that match \fISelection\fP criteria. All rows are displayed with the additional "selected" column (\fB-o selected\fP) -showing 1 if the row matches the Selection and 0 otherwise. For non-reporting +showing 1 if the row matches the \fISelection\fP and 0 otherwise. For non-reporting commands which process LVM entities, the selection can be used to match items to process. See \fBSELECTION CRITERIA\fP section of this man page for more information about the way the selection criteria are constructed. -.TP -.BR \-M ", " \-\-metadatatype " " \fIType -Specifies which type of on-disk metadata to use, such as \fIlvm1\fP -or \fIlvm2\fP, which can be abbreviated to \fI1\fP or \fI2\fP respectively. -The default (\fIlvm2\fP) can be changed by setting \fBformat\fP -in the \fBglobal\fP section of the config file. -.TP -.B \-\-ignorelockingfailure +. +.HP +.BR \-M | \-\-metadatatype +.IR Type +.br +Specifies which \fItype\fP of on-disk metadata to use, such as \fBlvm1\fP +or \fBlvm2\fP, which can be abbreviated to \fB1\fP or \fB2\fP respectively. +The default (\fBlvm2\fP) can be changed by setting \fBformat\fP +in the \fBglobal\fP section of the config file \fBlvm.conf\fP(5). +. +.HP +.BR \-\-ignorelockingfailure +.br This lets you proceed with read-only metadata operations such as \fBlvchange \-ay\fP and \fBvgchange \-ay\fP even if the locking module fails. One use for this is in a system init script if the lock directory is mounted read-only when the script runs. -.TP -.B \-\-ignoreskippedcluster +. +.HP +.BR \-\-ignoreskippedcluster +.br Use to avoid exiting with an non-zero status code if the command is run without clustered locking and some clustered Volume Groups have to be skipped over. -.TP -.B \-\-readonly +. +.HP +.BR \-\-readonly +.br Run the command in a special read-only mode which will read on-disk metadata without needing to take any locks. This can be used to peek inside metadata used by a virtual machine image while the virtual @@ -255,21 +357,28 @@ machine is running. It can also be used to peek inside the metadata of clustered Volume Groups when clustered locking is not configured or running. No attempt will be made to communicate with the device-mapper kernel driver, so -this option is unable to report whether or not Logical Volumes are +this option is unable to report whether or not Logical Volumes are actually in use. -.TP -.B \-\-foreign +. +.HP +.BR \-\-foreign +.br Cause the command to access foreign VGs, that would otherwise be skipped. It can be used to report or display a VG that is owned by another host. This option can cause a command to perform poorly because lvmetad caching is not used and metadata is read from disks. -.TP -.B \-\-shared +. +.HP +.BR \-\-shared +.br Cause the command to access shared VGs, that would otherwise be skipped when lvmlockd is not being used. It can be used to report or display a lockd VG without locking. -.TP -.B \-\-addtag \fITag +. +.HP +.BR \-\-addtag +.IR Tag +.br Add the tag \fITag\fP to a PV, VG or LV. Supply this argument multiple times to add more than one tag at once. A tag is a word that can be used to group LVM2 objects of the same type @@ -288,79 +397,113 @@ Only the new LVM2 metadata format supports tagging: objects using the LVM1 metadata format cannot be tagged because the on-disk format does not support it. Characters allowed in tags are: -.B A-Z a-z 0-9 _ + . - +.BR A - Z +.BR a - z +.BR 0 - 9 +.BR "_ + . -" and as of version 2.02.78 the following characters are also accepted: -.B / = ! : # & -.TP -.B \-\-deltag \fITag +.BR "/ = ! : # &" +.. +.HP +.BR \-\-deltag +.IR Tag +.br Delete the tag \fITag\fP from a PV, VG or LV, if it's present. Supply this argument multiple times to remove more than one tag at once. -.TP -.IR \fB\-\-alloc \ { anywhere | contiguous | cling | inherit | normal } +. +.HP +.BR \-\-alloc +.RB { anywhere | contiguous | cling | inherit | normal } +.br Selects the allocation policy when a command needs to allocate Physical Extents from the Volume Group. Each Volume Group and Logical Volume has an allocation policy defined. -The default for a Volume Group is \fInormal\fP which applies +The default for a Volume Group is \fBnormal\fP which applies common-sense rules such as not placing parallel stripes on the same -Physical Volume. The default for a Logical Volume is \fIinherit\fP +Physical Volume. The default for a Logical Volume is \fBinherit\fP which applies the same policy as for the Volume Group. These policies can be changed using \fBlvchange\fP(8) and \fBvgchange\fP(8) or overridden on the command line of any command that performs allocation. -The \fIcontiguous\fP policy requires that new Physical Extents be placed adjacent +The \fBcontiguous\fP policy requires that new Physical Extents be placed adjacent to existing Physical Extents. -The \fIcling\fP policy places new Physical Extents on the same Physical +The \fBcling\fP policy places new Physical Extents on the same Physical Volume as existing Physical Extents in the same stripe of the Logical Volume. If there are sufficient free Physical Extents to satisfy -an allocation request but \fInormal\fP doesn't use them, -\fIanywhere\fP will - even if that reduces performance by +an allocation request but \fBnormal\fP doesn't use them, +\fBanywhere\fP will - even if that reduces performance by placing two stripes on the same Physical Volume. -.TP -.IR \fB\-\-commandprofile " " \fIProfileName +. +.HP +.BR \-\-commandprofile +.IR ProfileName +.br Selects the command configuration profile to use when processing an LVM command. See also \fBlvm.conf\fP(5) for more information about \fBcommand profile config\fP and the way it fits with other LVM configuration methods. Using \fB\-\-commandprofile\fP option overrides any command profile specified via \fBLVM_COMMAND_PROFILE\fP environment variable. -.TP -.IR \fB\-\-metadataprofile " " \fIProfileName +. +.HP +.BR \-\-metadataprofile +.IR ProfileName +.br Selects the metadata configuration profile to use when processing an LVM command. When using metadata profile during Volume Group or Logical Volume creation, the metadata profile name is saved in metadata. When such Volume Group or Logical Volume is processed next time, the metadata profile is automatically applied -and the use of \-\-metadataprofile option is not necessary. See also +and the use of \fB\-\-metadataprofile\fP option is not necessary. See also \fBlvm.conf\fP(5) for more information about \fBmetadata profile config\fP and the way it fits with other LVM configuration methods. -.TP -.IR \fB\-\-profile " " \fIProfileName +. +.HP +.BR \-\-profile +.IR ProfileName +.br A short form of \fB\-\-metadataprofile\fP for \fBvgcreate\fP, \fBlvcreate\fP, \fBvgchange\fP and \fBlvchange\fP command and a short form of \fB\-\-commandprofile\fP for any other command (with the exception of \fBlvmconfig\fP command where the -\-\-profile has special meaning, see \fBlvmconfig\fP(8) for more information). -.TP -.IR \fB\-\-config " " \fIConfigurationString +\fB\-\-profile\fP has special meaning, see \fBlvmconfig\fP(8) for more information). +. +.HP +.BR \-\-config +.IR ConfigurationString +.br Uses the ConfigurationString as direct string representation of the configuration to override the existing configuration. The ConfigurationString is of exactly the same format as used in any LVM configuration file. See \fBlvm.conf\fP(5) for more information about \fBdirect config override on command line\fP and the way it fits with other LVM configuration methods. +. .SH VALID NAMES +. The valid characters for VG and LV names are: -.B a-z A-Z 0-9 + _ . - -.LP +.BR a - z +.BR A - Z +.BR 0 - 9 +.BR "+ _ . -" +.P VG and LV names cannot begin with a hyphen. -There are also various reserved names that are used internally by lvm that can not be used as LV or VG names. -A VG cannot be called anything that exists in /dev/ at the time of creation, nor can it be called '.' or '..'. -An LV cannot be called '.', '..', 'snapshot' or 'pvmove'. The LV name may also not contain any of the following -strings: '_cdata', '_cmeta', '_corig', '_mlog', '_mimage', '_pmspare', '_rimage', '_rlog', '_tdata' or '_tmeta'. -A directory bearing the name of each Volume Group is created under /dev when any of its Logical Volumes are activated. -Each active Logical Volume is accessible from this directory as a symbolic link leading to a device node. -Links or nodes in /dev/mapper are intended only for internal use and the precise format and escaping might change between releases and distributions. +There are also various reserved names that are used internally by lvm that can +not be used as LV or VG names. A VG cannot be called anything that exists in +\fI/dev/\fP at the time of creation, nor can it be called '.' or '..'. +An LV cannot be called '.', '..', 'snapshot' or 'pvmove'. +The LV name may also not contain any of the following strings: +'_cdata', '_cmeta', '_corig', '_mlog', '_mimage', '_pmspare', '_rimage', +'_rlog', '_tdata' or '_tmeta'. +A directory bearing the name of each Volume Group is created under +\fI/dev\fP when any of its Logical Volumes are activated. +Each active Logical Volume is accessible from this directory as a symbolic +link leading to a device node. +Links or nodes in \fI/dev/mapper\fP are intended only for internal use and +the precise format and escaping might change between releases and distributions. Other software and scripts should use the -/dev/VolumeGroupName/LogicalVolumeName format to reduce the chance of needing +\fI/dev/VolumeGroupName/LogicalVolumeName\fP format to reduce the chance of needing amendment when the software is updated. Should you need to process the node names in /dev/mapper, you may use \fBdmsetup splitname\fP to separate out the original VG, LV and internal layer names. +. .SH ALLOCATION +. When an operation needs to allocate Physical Extents for one or more Logical Volumes, the tools proceed as follows: @@ -370,7 +513,7 @@ the end of the command line, only unallocated Physical Extents within those ranges on the specified Physical Volumes are considered. Then they try each allocation policy in turn, starting with the strictest -policy (\fIcontiguous\fP) and ending with the allocation policy specified +policy (\fBcontiguous\fP) and ending with the allocation policy specified using \fB\-\-alloc\fP or set as the default for the particular Logical Volume or Volume Group concerned. For each policy, working from the lowest-numbered Logical Extent of the empty Logical Volume space that @@ -380,15 +523,15 @@ they move on to the next policy. The restrictions are as follows: -\fIContiguous\fP requires that the physical location of any Logical +\fBContiguous\fP requires that the physical location of any Logical Extent that is not the first Logical Extent of a Logical Volume is adjacent to the physical location of the Logical Extent immediately preceding it. -\fICling\fP requires that the Physical Volume used for any Logical +\fBCling\fP requires that the Physical Volume used for any Logical Extent to be added to an existing Logical Volume is already in use by at least one Logical Extent earlier in that Logical Volume. If the -configuration parameter allocation/cling_tag_list is defined, then two +configuration parameter \fBallocation/cling_tag_list\fP is defined, then two Physical Volumes are considered to match if any of the listed tags is present on both Physical Volumes. This allows groups of Physical Volumes with similar properties (such as their physical location) to be @@ -398,25 +541,25 @@ When a Logical Volume is striped or mirrored, the above restrictions are applied independently to each stripe or mirror image (leg) that needs space. -\fINormal\fP will not choose a Physical Extent that shares the same Physical +\fBNormal\fP will not choose a Physical Extent that shares the same Physical Volume as a Logical Extent already allocated to a parallel Logical -Volume (i.e. a different stripe or mirror image/leg) at the same offset +Volume (i.e. a different stripe or mirror image/leg) at the same offset within that parallel Logical Volume. When allocating a mirror log at the same time as Logical Volumes to hold the mirror data, Normal will first try to select different Physical Volumes for the log and the data. If that's not possible and the -allocation/mirror_logs_require_separate_pvs configuration parameter is -set to 0, it will then allow the log to share Physical Volume(s) with -part of the data. +.B allocation/mirror_logs_require_separate_pvs +configuration parameter is set to 0, it will then allow the log +to share Physical Volume(s) with part of the data. When allocating thin pool metadata, similar considerations to those of a mirror log in the last paragraph apply based on the value of the -allocation/thin_pool_metadata_require_separate_pvs configuration -parameter. +.B allocation/thin_pool_metadata_require_separate_pvs +configuration parameter. If you rely upon any layout behaviour beyond that documented here, be -aware that it might change in future versions of the code. +aware that it might change in future versions of the code. For example, if you supply on the command line two empty Physical Volumes that have an identical number of free Physical Extents available for @@ -431,87 +574,123 @@ discretion over the layout. To view the way the allocation process currently works in any specific case, read the debug logging output, for example by adding \fB\-vvvv\fP to a command. - +. .SH LOGICAL VOLUME TYPES +. Some logical volume types are simple to create and can be done with a single \fBlvcreate\fP(8) command. The linear and striped logical volume types are an example of this. Other logical volume types may require more than one command to create. The cache (\fBlvmcache\fP(7)) and thin provisioning (\fBlvmthin\fP(7)) types are examples of this. - +. .SH SELECTION CRITERIA -.LP +. The selection criteria are a set of \fBstatements\fP combined by \fBlogical and grouping operators\fP. The \fBstatement\fP consists of \fBcolumn\fP name for which a set of valid \fBvalues\fP is defined using \fBcomparison operators\fP. For complete list of column names (fields) that can be used in selection, see the output of \fB<lvm reporting command> -S help\fP. -.LP -\fBComparison operators\fP (cmp_op): -.IP -\fB=~\fP \(em Matching regular expression. -.IP -\fB!~\fP \(em Not matching regular expression. -.IP -\fB=\fP \(em Equal to. -.IP -\fB!=\fP \(em Not equal to. -.IP -\fB>=\fP \(em Greater than or equal to. -.IP -\fB>\fP \(em Greater than -.IP -\fB<=\fP \(em Less than or equal to. -.IP -\fB<\fP \(em Less than. -.LP -\fBBinary logical operators\fP (cmp_log): -.IP -\fB&&\fP \(em All fields must match -.IP -\fB,\fP \(em All fields must match -.IP -\fB||\fP \(em At least one field must match -.IP -\fB#\fP \(em At least one field must match -.LP -\fBUnary logical operators\fP: -.IP -\fB!\fP \(em Logical negation -.LP -\fBGrouping operators\fP: -.IP -\fB(\fP \(em Left parenthesis -.IP -\fB)\fP \(em Right parenthesis -.IP -\fB[\fP \(em List start -.IP -\fB]\fP \(em List end -.IP -\fB{\fP \(em List subset start -.IP -\fB}\fP \(em List subset end -.LP -\fBInformal grammar specification\fP: -.IP -.BR STATEMENT " = " column " cmp_op " VALUE " | " STATEMENT " log_op " STATEMENT " | " (STATEMENT) " | " !(STATEMENT) -.IP +.P +.SS Comparison operators \fR(cmp_op) +.sp +.PD 0 +.TP +.B =~ +Matching regular expression. +.TP +.B !~ +Not matching regular expression. +.TP +.B = +Equal to. +.TP +.B != +Not equal to. +.TP +.B >= +Greater than or equal to. +.TP +.B > +Greater than +.TP +.B <= +Less than or equal to. +.TP +.B < +Less than. +.PD +.P +.SS Binary logical operators \fR(cmp_log) +.sp +.PD 0 +.TP +.B && +All fields must match +.TP +.B , +All fields must match +.TP +.B || +At least one field must match +.TP +.B # +At least one field must match +.PD +.P +.SS Unary logical operators +.TP +.B ! +Logical negation +.P +.SS Grouping operators +.sp +.PD 0 +.TP +.B ( +Left parenthesis +.TP +.B ) +Right parenthesis +.TP +.B [ +List start +.TP +.B ] +List end +.TP +.B { +List subset start +.TP +.B } +List subset end +.PD +.SS Informal grammar specification +.HP +.BR STATEMENT " = " column " cmp_op " VALUE " | " \%STATEMENT " log_op " STATEMENT " | " \%(STATEMENT) " | " \%!(STATEMENT) +.br +.HP .BR VALUE " = " [VALUE " log_op " VALUE] .br -For list-based types: string list. Matches strictly. The log_op must always be of one type within the whole list value. -.IP +For list-based types: string list. Matches strictly. +The log_op must always be of one type within the whole list value. +.HP .BR VALUE " = " {VALUE " log_op " VALUE} .br -For list-based types: string list. Matches a subset. The log_op must always be of one type within the whole list value. -.IP +For list-based types: string list. Matches a subset. +The log_op must always be of one type within the whole list value. +.HP .BR VALUE " = " value .br -For scalar types: number (integer), size (floating point number with size unit suffix), percent (floating point number with or without % suffix), string. - +For scalar types: number (integer), size (floating point number +with size unit suffix), percent (floating point number with or +without % suffix), string. +. .SH DIAGNOSTICS +. All tools return a status code of zero on success or non-zero on failure. +. .SH ENVIRONMENT VARIABLES +. .TP .B HOME Directory containing \fI.lvm_history\fP if the internal readline @@ -523,7 +702,7 @@ is overriden by direct use of \fB\-\-commandprofile\fP command line option. .TP .B LVM_SYSTEM_DIR Directory containing \fBlvm.conf\fP(5) and other LVM system files. -Defaults to "#DEFAULT_SYS_DIR#". +Defaults to "\fI#DEFAULT_SYS_DIR#\fP". .TP .B LVM_SUPPRESS_FD_WARNINGS Suppress warnings about unexpected file descriptors passed into LVM. @@ -554,9 +733,10 @@ separate file. The status anticipated when the process exits. Use ">N" to match any status greater than N. If the actual exit status matches and a log file got produced, it is deleted. - -LVM_LOG_FILE_EPOCH and LVM_EXPECTED_EXIT_STATUS together allow -automated test scripts to discard uninteresting log data. +.B LVM_LOG_FILE_EPOCH +and +.B LVM_EXPECTED_EXIT_STATUS +together allow automated test scripts to discard uninteresting log data. .TP .B LVM_SUPPRESS_LOCKING_FAILURE_MESSAGES Used to suppress warning messages when the configured locking is known @@ -568,11 +748,15 @@ Abort processing if the code detects a non-fatal internal error. .B DM_DISABLE_UDEV Avoid interaction with udev. LVM will manage the relevant nodes in /dev directly. +. .SH FILES +. .I #DEFAULT_SYS_DIR#/lvm.conf .br .I $HOME/.lvm_history +. .SH SEE ALSO +. .BR lvm.conf (5), .BR lvmcache (7), .BR lvmthin (7),

1 0

master - man: update dmsetup and dmstats pages
by Zdenek Kabelac 01 Oct '15

01 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=867a36b4197e2abe… Commit: 867a36b4197e2abe6efd3b16e229590eedc566a3 Parent: a139275eca4fadca4a456fd27e8b2ba30d7d2c02 Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Wed Sep 23 11:25:00 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 1 15:03:12 2015 +0200 man: update dmsetup and dmstats pages Try to provide properly rendered pages no just with plain 'man' but also for: man -Tps man -Thtml man2html --- man/dmsetup.8.in | 1039 +++++++++++++++++++++++++++++++++--------------------- man/dmstats.8.in | 890 ++++++++++++++++++++++++---------------------- 2 files changed, 1099 insertions(+), 830 deletions(-) diff --git a/man/dmsetup.8.in b/man/dmsetup.8.in index eff805a..183f4fc 100644 --- a/man/dmsetup.8.in +++ b/man/dmsetup.8.in @@ -1,171 +1,378 @@ .TH DMSETUP 8 "Apr 06 2006" "Linux" "MAINTENANCE COMMANDS" +. .SH NAME +. dmsetup \(em low level logical volume management +. .SH SYNOPSIS +. +.\".nh .ad l .PD 0 +.HP 9 +.B dmsetup +.de CMD_CLEAR +. BR clear +. IR device_name +.. +.CMD_CLEAR +. +.HP +.B dmsetup +.de CMD_CREATE +. ad l +. BR create +. IR device_name +. RB [ -u | \-\-uuid +. IR uuid ] +. RB \%[ \-\-addnodeoncreate | \-\-addnodeonresume ] +. RB \%[ \-n | \-\-notable | \-\-table +. RI \%{ table | table_file }] +. RB [ \-\-readahead +. RB \%{[ + ] \fIsectors | auto | none }] +. ad b +.. +.CMD_CREATE +. +.HP +.B dmsetup +.de CMD_DEPS +. ad l +. BR deps +. RB [ \-o +. IR options ] +. RI [ device_name ] +. ad b +.. +.CMD_DEPS +. +.HP +.B dmsetup +.de CMD_HELP +. BR help +. RB [ \-c | \-C | \-\-columns ] +.. +.CMD_HELP +. +.HP +.B dmsetup +.de CMD_INFO +. BR info +. RI [ device_name ] +.. +.CMD_INFO +. +.HP +.B dmsetup +.de CMD_INFOLONG +. ad l +. BR info +. BR \-c | \-C | \-\-columns +. RB [ \-\-count +. IR count ] +. RB [ \-\-interval +. IR seconds ] +. RB \%[ \-\-nameprefixes ] +. RB \%[ \-\-noheadings ] +. RB [ \-o +. IR fields ] +. RB [ \-O | \-\-sort +. IR sort_fields ] +. RB [ \-\-separator +. IR separator ] +. RI [ device_name ] +. ad b +.. +.CMD_INFOLONG +. +.HP +.B dmsetup +.de CMD_LOAD +. ad l +. BR load +. IR device_name +. RB [ \-\-table +. RI { table | table_file }] +. ad b +.. +.CMD_LOAD +. +.HP +.B dmsetup +.de CMD_LS +. ad l +. BR ls +. RB [ \-\-target +. IR target_type ] +. RB [ \-\-exec +. IR command ] +. RB [ \-\-tree ] +. RB [ \-o +. IR options ] +. ad b +.. +.CMD_LS +. +.HP +.B dmsetup +.de CMD_MANGLE +. BR mangle +. RI [ device_name ] +.. +.CMD_MANGLE +. +.HP +.B dmsetup +.de CMD_MESSAGE +. BR message +. IR device_name +. IR sector +. IR message +.. +.CMD_MESSAGE +. +.HP +.B dmsetup +.de CMD_MKNODES +. BR mknodes +. RI [ device_name ] +.. +.CMD_MKNODES +. +.HP +.B dmsetup +.de CMD_RELOAD +. ad l +. BR reload +. IR device_name +. RB [ \-\-table +. RI { table | table_file }] +. ad b +.. +.CMD_RELOAD +. +.HP +.B dmsetup +.de CMD_REMOVE +. ad l +. BR remove +. RB [ \-f | \-\-force ] +. RB [ \-\-retry ] +. RB [ \-\-deferred ] +. IR device_name +. ad b +.. +.CMD_REMOVE +. +.HP +.B dmsetup +.de CMD_REMOVE_ALL +. BR remove_all +. RB [ \-f | \-\-force ] +. RB [ \-\-deferred ] +.. +.CMD_REMOVE_ALL +. +.HP +.B dmsetup +.de CMD_RENAME +. BR rename +. IR device_name +. IR new_name +.. +.CMD_RENAME +. +.HP +.B dmsetup +.de CMD_RENAME_UUID +. BR rename +. IR device_name +. BR \-\-setuuid +. IR uuid +.. +.CMD_RENAME_UUID +. +.HP +.B dmsetup +.de CMD_RESUME +. ad l +. BR resume +. IR device_name +. RB [ \-\-addnodeoncreate | \-\-addnodeonresume ] +. RB [ \-\-noflush ] +. RB [ \-\-nolockfs ] +. RB \%[ \-\-readahead +. RB \%{[ + ] \fIsectors | auto | none }] +. ad b +.. +.CMD_RESUME +. +.HP +.B dmsetup +.de CMD_SETGEOMETRY +. ad l +. BR setgeometry +. IR device_name +. IR cyl +. IR head +. IR sect +. IR start +. ad b +.. +.CMD_SETGEOMETRY +. +.HP +.B dmsetup +.de CMD_SPLITNAME +. BR splitname +. IR device_name +. RI [ subsystem ] +.. +.CMD_SPLITNAME +. +.HP +.B dmsetup +.de CMD_STATS +. BR stats +. IR command +. RI [ options ] +.. +.CMD_STATS +. +.HP +.B dmsetup +.de CMD_STATUS +. ad l +. BR status +. RB [ \-\-target +. IR target_type ] +. RB [ \-\-noflush ] +. RI [ device_name ] +. ad b +.. +.CMD_STATUS +. +.HP +.B dmsetup +.de CMD_SUSPEND +. ad l +. BR suspend +. RB [ \-\-nolockfs ] +. RB [ \-\-noflush ] +. IR device_name +. ad b +.. +.CMD_SUSPEND +. +.HP +.B dmsetup +.de CMD_TABLE +. ad l +. BR table +. RB [ \-\-target +. IR target_type ] +. RB [ \-\-showkeys ] +. RI [ device_name ] +. ad b +.. +.CMD_TABLE +. +.HP +.B dmsetup +.de CMD_TARGETS +. BR targets +.. +.CMD_TARGETS +. +.HP +.B dmsetup +.de CMD_UDEVCOMPLETE +. BR udevcomplete +. IR cookie +.. +.CMD_UDEVCOMPLETE +. +.HP +.B dmsetup +.de CMD_UDEVCOMPLETE_ALL +. BR udevcomplete_all +. RI [ age_in_minutes ] +.. +.CMD_UDEVCOMPLETE_ALL +. +.HP +.B dmsetup +.de CMD_UDEVCOOKIES +. BR udevcookie +.. +.CMD_UDEVCOOKIES +. +.HP +.B dmsetup +.de CMD_UDEVCTEATECOOKIE +. BR udevcreatecookie +.. +.CMD_UDEVCTEATECOOKIE +. +.HP +.B dmsetup +.de CMD_UDEVFLAGS +. BR udevflags +. IR cookie +.. +.CMD_UDEVFLAGS +. +.HP +.B dmsetup +.de CMD_UDEVRELEASECOOKIE +. BR udevreleasecookie +. RI [ cookie ] +.. +.CMD_UDEVRELEASECOOKIE +. +.HP +.B dmsetup +.de CMD_VERSION +. BR version +.. +.CMD_VERSION +. +.HP +.B dmsetup +.de CMD_WAIT +. ad l +. BR wait +. RB [ \-\-noflush ] +. IR device_name +. RI [ event_nr ] +. ad b +.. +.CMD_WAIT +. +.HP +.B dmsetup +.de CMD_WIPE_TABLE +. ad l +. BR wipe_table +. IR device_name +. RB [ \-f | \-\-force ] +. RB [ \-\-noflush ] +. RB [ \-\-nolockfs ] +. ad b +.. +.CMD_WIPE_TABLE +.PD +.P .HP -.B dmsetup clear -.I device_name -.HP -.B dmsetup create -.I device_name -.RB [ \-\-addnodeoncreate | \-\-addnodeonresume ] -.RB [ \-\-readahead -.RB {[ + ] \fIsectors | auto | none }] -.RB [ \-u -.IR uuid ] -.RB [ \-n | \-\-notable | \-\-table -.RI { table | table_file }] -.HP -.B dmsetup deps -.RB [ \-o -.IR options ] -.RI [ device_name ] -.HP -.B dmsetup help -.RB [ \-c | \-C | \-\-columns ] -.HP -.B dmsetup info -.RI [ device_name ] -.HP -.B dmsetup info -.BR \-c | \-C | \-\-columns -.RB [ \-\-count -.IR count ] -.RB [ \-\-interval -.IR seconds ] -.RB [ \-\-nameprefixes ] -.RB [ \-\-noheadings ] -.RB [ \-o -.IR fields ] -.RB [ \-O | \-\-sort -.IR sort_fields ] -.RB [ \-S | \-\-select -.IR selection ] -.RB [ \-\-separator -.IR separator ] -.RI [ device_name ] -.HP -.B dmsetup load -.I device_name -.RB [ \-\-table -.RI { table | table_file }] -.HP -.B dmsetup ls -.RB [ \-\-target -.IR target_type ] -.RB [ \-\-exec -.IR command ] -.RB [ \-\-tree ] -.RB [ \-o -.IR options ] -.HP -.B dmsetup message -.I device_name sector message -.HP -.B dmsetup mknodes -.RI [ device_name ] -.HP -.B dmsetup mangle -.RI [ device_name ] -.HP -.B dmsetup reload -.I device_name -.RB [ \-\-table -.RI { table | table_file }] -.HP -.B dmsetup remove -.RB [ \-f | \-\-force ] -.RB [ \-\-retry ] -.RB [ \-\-deferred ] -.I device_name -.HP -.B dmsetup remove_all -.RB [ \-f | \-\-force ] -.RB [ \-\-deferred ] -.HP -.B dmsetup rename -.I device_name new_name -.HP -.B dmsetup rename -.I device_name -.B \-\-setuuid -.I uuid -.HP -.B dmsetup resume -.I device_name -.RB [ \-\-addnodeoncreate | \-\-addnodeonresume ] -.RB [ \-\-readahead -.RB {[ + ] \fIsectors | auto | none }] -.HP -.B dmsetup setgeometry -.I device_name cyl head sect start -.HP -.B dmsetup splitname -.I device_name -.RI [ subsystem ] -.HP -.B dmsetup stats -.I command -.RI [ options ] -.HP -.B dmsetup status -.RB [ \-\-target -.IR target_type ] -.RB [ \-\-noflush ] -.RI [ device_name ] -.HP -.B dmsetup suspend -.RB [ \-\-nolockfs ] -.RB [ \-\-noflush ] -.I device_name -.HP -.B dmsetup table -.RB [ \-\-target -.IR target_type ] -.RB [ \-\-showkeys ] -.RI [ device_name ] -.HP -.B dmsetup targets -.HP -.B dmsetup udevcomplete -.I cookie -.HP -.B dmsetup udevcomplete_all -.RI [ age_in_minutes ] -.HP -.B dmsetup udevcookies -.HP -.B dmsetup udevcreatecookie -.HP -.B dmsetup udevflags -.I cookie -.HP -.B dmsetup udevreleasecookie -.RI [ cookie ] -.HP -.B dmsetup version -.HP -.B dmsetup wait -.RB [ \-\-noflush ] -.I device_name -.RI [ event_nr ] -.HP -.B dmsetup wipe_table -.I device_name -.RB [ \-f | \-\-force ] -.RB [ \-\-noflush ] -.RB [ \-\-nolockfs ] -.HP -.B devmap_name -.I major minor +.PD 0 +.B devmap_name \fImajor minor .HP -.B devmap_name -.I major:minor -.ad b +.B devmap_name \fImajor:minor .PD +.ad b +. .SH DESCRIPTION +. dmsetup manages logical devices that use the device-mapper driver. Devices are created by loading a table that specifies a target for each sector (512 bytes) in the logical device. @@ -173,166 +380,235 @@ each sector (512 bytes) in the logical device. The first argument to dmsetup is a command. The second argument is the logical device name or uuid. -Invoking the command as \fBdevmap_name\fP is equivalent to -.br -\fBdmsetup info \-c \-\-noheadings \-j \fImajor\fB \-m \fIminor\fP. +Invoking the dmsetup tool as \fBdevmap_name\fP +(which is not normally distributed and is supported +only for historical reasons) is equivalent to +.BI \%dmsetup\ info\ \-c\ \-\-noheadings\ \-j \ major\ \-m \ minor \c +\fR. +.\" dot above here fixes -Thtml rendering for next HP option +. .SH OPTIONS -.TP -.B \-\-addnodeoncreate -Ensure /dev/mapper node exists after dmsetup create. -.TP -.B \-\-addnodeonresume -Ensure /dev/mapper node exists after dmsetup resume (default with udev). -.TP -.B \-\-checks +. +.HP +.BR \-\-addnodeoncreate +.br +Ensure \fI/dev/mapper\fP node exists after \fBdmsetup create\fP. +. +.HP +.BR \-\-addnodeonresume +.br +Ensure \fI/dev/mapper\fP node exists after \fBdmsetup resume\fP (default with udev). +. +.HP +.BR \-\-checks +.br Perform additional checks on the operations requested and report potential problems. Useful when debugging scripts. In some cases these checks may slow down operations noticeably. -.TP +. +.HP .BR \-c | \-C | \-\-columns +.br Display output in columns rather than as Field: Value lines. -.TP -.B \-\-count \fIcount +. +.HP +.BR \-\-count +.IR count +.br Specify the number of times to repeat a report. Set this to zero continue until interrupted. The default interval is one second. -.TP +. +.HP .BR \-f | \-\-force +.br Try harder to complete operation. -.TP +. +.HP .BR \-h | \-\-help +.br Outputs a summary of the commands available, optionally including the list of report fields (synonym with \fBhelp\fP command). -.TP -.B \-\-inactive +. +.HP +.BR \-\-inactive +.br When returning any table information from the kernel report on the inactive table instead of the live table. Requires kernel driver version 4.16.0 or above. -.TP -.B \-\-interval \fIseconds +. +.HP +.BR \-\-interval +.IR seconds +.br Specify the interval in seconds between successive iterations for repeating reports. If \fB\-\-interval\fP is specified but \fB\-\-count\fP is not, reports will continue to repeat until interrupted. The default interval is one second. -.TP -.BR \-\-manglename \ { none | hex | auto } +. +.HP +.BR \-\-manglename +.RB { auto | hex | none } +.br Mangle any character not on a whitelist using mangling_mode when processing device-mapper device names and UUIDs. The names and UUIDs are mangled on input and unmangled on output where the mangling mode is one of: -\fBnone\fP (no mangling), -\fBhex\fP (always do the mangling) and \fBauto\fP (only do the mangling if not mangled yet, do nothing -if already mangled, error on mixed) +if already mangled, error on mixed), +\fBhex\fP (always do the mangling) and +\fBnone\fP (no mangling). Default mode is \fB#DEFAULT_MANGLING#\fP. Character whitelist: 0-9, A-Z, a-z, #+-.:=@_. This whitelist is also supported by udev. Any character not on a whitelist is replaced with its hex value (two digits) prefixed by \\x. Mangling mode could be also set through -.B DM_DEFAULT_NAME_MANGLING_MODE +\fBDM_DEFAULT_NAME_MANGLING_MODE\fP environment variable. -.TP -.BR \-j | \-\-major\ \fImajor +. +.HP +.BR \-j | \-\-major +.IR major +.br Specify the major number. -.TP -.BR \-m | \-\-minor\ \fIminor +. +.HP +.BR \-m | \-\-minor +.IR minor +.br Specify the minor number. -.TP +. +.HP .BR \-n | \-\-notable +.br When creating a device, don't load any table. -.TP +. +.HP .BR \-\-nameprefixes +.br Add a "DM_" prefix plus the field name to the output. Useful with \fB\-\-noheadings\fP to produce a list of field=value pairs that can be used to set environment variables (for example, in .BR udev (7) rules). -.TP +. +.HP .BR \-\-noheadings Suppress the headings line when using columnar output. -.TP -.B \-\-noflush +. +.HP +.BR \-\-noflush Do not flush outstading I/O when suspending a device, or do not commit thin-pool metadata when obtaining thin-pool status. -.TP -.B \-\-nolockfs +. +.HP +.BR \-\-nolockfs +.br Do not attempt to synchronize filesystem eg, when suspending a device. -.TP -.B \-\-noopencount +. +.HP +.BR \-\-noopencount +.br Tell the kernel not to supply the open reference count for the device. -.TP -.B \-\-noudevrules +. +.HP +.BR \-\-noudevrules +.br Do not allow udev to manage nodes for devices in device-mapper directory. -.TP -.B \-\-noudevsync +. +.HP +.BR \-\-noudevsync +.br Do not synchronise with udev when creating, renaming or removing devices. -.TP +. +.HP .BR \-o | \-\-options +.IR options +.br Specify which fields to display. -.TP -.BR \-\-readahead \ {[ + ] \fIsectors | auto | none } +. +.HP +.BR \-\-readahead +.RB {[ + ] \fIsectors | auto | none } +.br Specify read ahead size in units of sectors. The default value is \fBauto\fP which allows the kernel to choose a suitable value automatically. The \fB+\fP prefix lets you specify a minimum value which will not be used if it is smaller than the value chosen by the kernel. The value \fBnone\fP is equivalent to specifying zero. -.TP +. +.HP .BR \-r | \-\-readonly +.br Set the table being loaded read-only. -.TP -.BR \-S | \-\-select \ \fIselection +. +.HP +.BR \-S | \-\-select +.IR selection +.br Display only rows that match \fIselection\fP criteria. All rows are displayed with the additional "selected" column (\fB-o selected\fP) showing 1 if the row matches the \fIselection\fP and 0 otherwise. The selection criteria are defined by specifying column names and their valid values while making use of supported comparison operators. As a quick help and to see full list of column names that can be used in selection and the set of supported -selection operators, check the output of \fBdmsetup info -c -S help\fP +selection operators, check the output of \fBdmsetup\ info\ -c\ -S\ help\fP command. -.TP -.IR \fB\-\-table \ table +. +.HP +.BR \-\-table +.IR table +.br Specify a one-line table directly on the command line. See below for more information on the table format. -.TP -.B \-\-udevcookie \fIcookie +. +.HP +.BR \-\-udevcookie +.IR cookie +.br Use cookie for udev synchronisation. Note: Same cookie should be used for same type of operations i.e. creation of multiple different devices. It's not adviced to combine different operations on the single device. -.TP +. +.HP .BR \-u | \-\-uuid +.br Specify the \fIuuid\fP. -.TP +. +.HP .BR \-y | \-\-yes +.br Answer yes to all prompts automatically. -.TP -.BR \-v | \-\-verbose \ [ \-v | \-\-verbose ] +. +.HP +.BR \-v | \-\-verbose +.RB [ \-v | \-\-verbose ] +.br Produce additional output. -.TP -.B \-\-verifyudev +. +.HP +.BR \-\-verifyudev +.br If udev synchronisation is enabled, verify that udev operations get performed correctly and try to fix up the device nodes afterwards if not. -.TP -.B \-\-version +. +.HP +.BR \-\-version +.br Display the library and kernel driver version. .br +. .SH COMMANDS +. .HP -.B clear -.I device_name +.CMD_CLEAR .br Destroys the table in the inactive table slot for device_name. +. .HP -.B create -.I device_name -.RB [ \-u -.IR uuid ] -.RB [ \-\-addnodeoncreate | \-\-addnodeonresume ] -.RB [ \-n | \-\-notable | \-\-table -.RI { table | table_file }] -.RB [ \-\-readahead -.RB {[ + ] \fIsectors | auto | none }] +.CMD_CREATE .br Creates a device with the given name. If \fItable\fP or \fItable_file\fP is supplied, the table is loaded and made live. @@ -340,28 +616,26 @@ Otherwise a table is read from standard input unless \fB\-\-notable\fP is used. The optional \fIuuid\fP can be used in place of device_name in subsequent dmsetup commands. If successful the device will appear in table and for live -device the node /dev/mapper/\fIdevice_name\fP is created. +device the node \fI/dev/mapper/device_name\fP is created. See below for more information on the table format. +. .HP -.B deps -.RB [ \-o -.IR options ] -.RI [ device_name ] +.CMD_DEPS .br Outputs a list of devices referenced by the live table for the specified device. Device names on output can be customised by following \fIoptions\fP: \fBdevno\fP (major and minor pair, used by default), \fBblkdevname\fP (block device name), \fBdevname\fP (map name for device-mapper devices, equal to blkdevname otherwise). +. .HP -.B help -.RB [ \-c | \-C | \-\-columns ] +.CMD_HELP .br Outputs a summary of the commands available, optionally including the list of report fields. +. .HP -.B info -.RI [ device_name ] +.CMD_INFO .br Outputs some brief information about the device in the form: .RS @@ -376,45 +650,38 @@ Outputs some brief information about the device in the form: .RE .RE .HP -.B info -.BR \-c | \-C | \-\-columns -.RB [ \-\-count -.IR count ] -.RB [ \-\-interval -.IR seconds ] -.RB [ \-\-nameprefixes ] -.RB [ \-\-noheadings ] -.RB [ \-o -.IR fields ] -.RB [ \-O | \-\-sort -.IR sort_fields ] -.RB [ \-\-separator -.IR separator ] -.RI [ device_name ] +.CMD_INFOLONG .br Output you can customise. Fields are comma-separated and chosen from the following list: -name, major, minor, attr, open, segments, events, uuid. -Attributes are: (L)ive, (I)nactive, (s)uspended, (r)ead-only, read-(w)rite. +.BR name , +.BR major , +.BR minor , +.BR attr , +.BR open , +.BR segments , +.BR events , +.BR uuid . +Attributes are: +.RI ( L )ive, +.RI ( I )nactive, +.RI ( s )uspended, +.RI ( r )ead-only, +.RI read-( w )rite. Precede the list with '\fB+\fP' to append to the default selection of columns instead of replacing it. -Precede any sort_field with '\fB-\fP' for a reverse sort on that column. +Precede any sort field with '\fB-\fP' for a reverse sort on that column. +. .HP -.B ls -.RB [ \-\-target -.IR target_type ] -.RB [ \-\-exec -.IR command ] -.RB [ \-\-tree ] -.RB [ \-o -.IR options ] +.CMD_LS .br List device names. Optionally only list devices that have at least one target of the specified type. Optionally execute a command for each device. The device name is appended to the supplied command. -Device names on output can be customised by following options: devno (major -and minor pair, used by default), blkdevname (block device name), -devname (map name for device-mapper devices, equal to blkdevname otherwise). +Device names on output can be customised by following options: +\fBdevno\fP (major and minor pair, used by default), +\fBblkdevname\fP (block device name), +\fBdevname\fP (map name for device-mapper devices, equal to blkdevname otherwise). \fB\-\-tree\fP displays dependencies between devices as a tree. It accepts a comma-separate list of \fIoptions\fP. Some specify the information displayed against each node: @@ -424,17 +691,16 @@ Some specify the information displayed against each node: Others specify how the tree is displayed: .BR ascii ", " utf ", " vt100 ; .BR compact ", " inverted ", " notrunc . +. .HP -.BR load | reload -.I device_name -.RB [ \-\-table -.RI { table | table_file }] +.BR load | \c +.CMD_RELOAD .br Loads \fItable\fP or \fItable_file\fP into the inactive table slot for device_name. If neither is supplied, reads a table from standard input. +. .HP -.B mangle -.RI [ device_name ] +.CMD_MANGLE .br Ensure existing device-mapper \fIdevice_name\fP and UUID is in the correct mangled form containing only whitelisted characters (supported by udev) and do @@ -444,25 +710,22 @@ names and not for device UUIDs because the kernel does not allow changing the UUID of active devices. Any incorrect UUIDs are reported only and they must be manually corrected by deactivating the device first and then reactivating it with proper mangling mode used (see also \fB\-\-manglename\fP). +. .HP -.B message -.I device_name sector message +.CMD_MESSAGE .br Send message to target. If sector not needed use 0. +. .HP -.B mknodes -.RI [ device_name ] +.CMD_MKNODES .br -Ensure that the node in /dev/mapper for device_name is correct. -If no device_name is supplied, ensure that all nodes in /dev/mapper +Ensure that the node in \fI/dev/mapper\fP for \fIdevice_name\fP is correct. +If no device_name is supplied, ensure that all nodes in \fI/dev/mapper\fP correspond to mapped devices currently loaded by the device-mapper kernel driver, adding, changing or removing nodes as necessary. +. .HP -.B remove -.RB [ \-f | \-\-force ] -.RB [ \-\-retry ] -.RB [ \-\-deferred ] -.I device_name +.CMD_REMOVE .br Removes a device. It will no longer be visible to dmsetup. Open devices cannot be removed, but adding \fB\-\-force\fP will replace the table with one @@ -477,10 +740,9 @@ Do NOT combine \fB\-\-force\fP and \fB\-\-udevcookie\fP, as udev may start to process udev rules in the middle of error target replacement and result in nondeterministic result. +. .HP -.B remove_all -.RB [ \-f | \-\-force ] -.RB [ \-\-deferred ] +.CMD_REMOVE_ALL .br Attempts to remove all device definitions i.e. reset the driver. This also runs \fBmknodes\fP afterwards. Use with care! Open devices cannot be removed, but @@ -489,42 +751,34 @@ adding \fB\-\-force\fP will replace the table with one that fails all I/O. will be removed when the last user closes it. The deferred removal feature is supported since version 4.27.0 of the device-mapper driver available in upstream kernel version 3.13. +. .HP -.B rename -.I device_name new_name +.CMD_RENAME .br Renames a device. +. .HP -.B rename -.I device_name -.B \-\-setuuid -.I uuid +.CMD_RENAME_UUID .br Sets the uuid of a device that was created without a uuid. After a uuid has been set it cannot be changed. +. .HP -.B resume -.I device_name -.RB [ \-\-addnodeoncreate | \-\-addnodeonresume ] -.RB [ \-\-nolockfs ] -.RB [ \-\-noflush ] -.RB [ \-\-readahead -.RB {[ + ] \fIsectors | auto | none }] +.CMD_RESUME .br Un-suspends a device. If an inactive table has been loaded, it becomes live. Postponed I/O then gets re-queued for processing. +. .HP -.B setgeometry -.I device_name cyl head sect start +.CMD_SETGEOMETRY .br Sets the device geometry to C/H/S. +. .HP -.B splitname -.I device_name -.RI [ subsystem ] +.CMD_SPLITNAME .br -Splits given device name into subsystem constituents. +Splits given \fIdevice name\fP into \fIsubsystem\fP constituents. The default subsystem is LVM. LVM currently generates device names by concatenating the names of the Volume Group, Logical Volume and any internal Layer with a hyphen as separator. @@ -533,30 +787,22 @@ The precise encoding might change without notice in any future release, so we recommend you always decode using the current version of this command. .HP -.B stats -.I command -.RI [ options ] +.BR stats " " \fIcommand " [" \fIoptions ] .br Manages IO statistics regions for devices. -See +See .BR dmstats (8) for more details. .HP -.B status -.RB [ \-\-target -.IR target_type ] -.RB [ \-\-noflush ] -.RI [ device_name ] +.CMD_STATS .br Outputs status information for each of the device's targets. With \fB\-\-target\fP, only information relating to the specified target type any is displayed. With \fB\-\-noflush\fP, the thin target (from version 1.3.0) doesn't commit any outstanding changes to disk before reporting its statistics. + .HP -.B suspend -.RB [ \-\-nolockfs ] -.RB [ \-\-noflush ] -.I device_name +.CMD_SUSPEND .br Suspends a device. Any I/O that has already been mapped by the device but has not yet completed will be flushed. Any further I/O to that @@ -566,12 +812,9 @@ an attempt will be made to sync it first unless \fB\-\-nolockfs\fP is specified. Some targets such as recent (October 2006) versions of multipath may support the \fB\-\-noflush\fP option. This lets outstanding I/O that has not yet reached the device to remain unflushed. +. .HP -.B table -.RB [ \-\-target -.IR target_type ] -.RB [ \-\-showkeys ] -.RI [ device_name ] +.CMD_TABLE .br Outputs the current table for the device in a format that can be fed back in using the create or load commands. @@ -579,28 +822,31 @@ With \fB\-\-target\fP, only information relating to the specified target type is displayed. Encryption keys are suppressed in the table output for the crypt target unless the \fB\-\-showkeys\fP parameter is supplied. +. .HP -.B targets +.CMD_TARGETS .br Displays the names and versions of the currently-loaded targets. +. .HP -.B udevcomplete -.I cookie +.CMD_UDEVCOMPLETE .br Wake any processes that are waiting for udev to complete processing the specified cookie. +. .HP -.B udevcomplete_all -.RI [ age_in_minutes ] +.CMD_UDEVCOMPLETE_ALL .br Remove all cookies older than the specified number of minutes. Any process waiting on a cookie will be resumed immediately. +. .HP -.B udevcookies +.CMD_UDEVCOOKIES .br List all existing cookies. Cookies are system-wide semaphores with keys prefixed by two predefined bytes (0x0D4D). +. .HP -.B udevcreatecookie +.CMD_UDEVCREATECOOKIE .br Creates a new cookie to synchronize actions with udev processing. The output is a cookie value. Normally we don't need to create cookies since @@ -612,33 +858,32 @@ of the dmsetup process as \fBDM_UDEV_COOKIE\fP variable and it will be used auto with all subsequent commands until it is unset. Invoking this command will create system-wide semaphore that needs to be cleaned up explicitly by calling udevreleasecookie command. +. .HP -.B udevflags -.I cookie +.CMD_UDEVFLAGS .br -Parses given cookie value and extracts any udev control flags encoded. +Parses given \fIcookie\fP value and extracts any udev control flags encoded. The output is in environment key format that is suitable for use in udev rules. If the flag has its symbolic name assigned then the output is -DM_UDEV_FLAG_<flag_name>='1', DM_UDEV_FLAG<flag_position>='1' otherwise. +DM_UDEV_FLAG_<flag_name> = '1', DM_UDEV_FLAG<flag_position> = '1' otherwise. Subsystem udev flags don't have symbolic names assigned and these ones are -always reported as DM_SUBSYSTEM_UDEV_FLAG<flag_position>='1'. There are +always reported as DM_SUBSYSTEM_UDEV_FLAG<flag_position> = '1'. There are 16 udev flags altogether. +. .HP -.B udevreleasecookie -.RI [ cookie ] +.CMD_UDEVRELEASECOOKIE .br Waits for all pending udev processing bound to given cookie value and clean up the cookie with underlying semaphore. If the cookie is not given directly, the command will try to use a value defined by \fBDM_UDEV_COOKIE\fP environment variable. +. .HP -.B version +.CMD_VERSION .br Outputs version information. +. .HP -.B wait -.RB [ \-\-noflush ] -.I device_name -.RI [ event_nr ] +.CMD_WAIT .br Sleeps until the event counter for device_name exceeds event_nr. Use \fB\-v\fP to see the event number returned. @@ -646,39 +891,30 @@ To wait until the next event is triggered, use \fBinfo\fP to find the last event number. With \fB\-\-noflush\fP, the thin target (from version 1.3.0) doesn't commit any outstanding changes to disk before reporting its statistics. +. .HP -.B wipe_table -.I device_name -.RB [ \-f | \-\-force ] -.RB [ \-\-noflush ] -.RB [ \-\-nolockfs ] +.CMD_WIPE_TABLE .br Wait for any I/O in-flight through the device to complete, then replace the table with a new table that fails any new I/O sent to the device. If successful, this should release any devices held open by the device's table(s). +. .SH TABLE FORMAT +. Each line of the table specifies a single target and is of the form: -.P +.sp .I logical_start_sector num_sectors .B target_type .I target_args -.P -Simple target types and \fItarget_args\fP include: -.HP -.PD 0 -.B linear -.I destination_device -.I start_sector -.br +.sp +Simple target types and target args include: +. +.TP +.B linear \fIdestination_device start_sector The traditional linear mapping. -.HP -.B striped -.I num_stripes -.I chunk_size -.RI [ destination -.IR start_sector ]... -.br +.TP +.B striped \fInum_stripes chunk_size \fR[\fIdestination start_sector\fR]... Creates a striped area. .br e.g. striped 2 32 /dev/hda1 0 /dev/hdb1 0 @@ -692,93 +928,78 @@ will map the first chunk (16k) as follows: etc. .RE .RE -.HP +.TP .B error -.br Errors any I/O that goes to this area. Useful for testing or for creating devices with holes in them. -.HP +.TP .B zero -.br Returns blocks of zeroes on reads. Any data written is discarded silently. -This is a block-device equivalent of the /dev/zero character-device data sink -described in \fBnull\fP(4). -.PD -.HP +This is a block-device equivalent of the \fI/dev/zero\fP +character-device data sink described in \fBnull\fP(4). +.P More complex targets include: -.HP -.PD 0 +.TP .B cache -.br Improves performance of a block device (eg, a spindle) by dynamically migrating some of its data to a faster smaller device (eg, an SSD). -.HP +.TP .B crypt -.br Transparent encryption of block devices using the kernel crypto API. -.HP +.TP .B delay -.br Delays reads and/or writes to different devices. Useful for testing. -.HP +.TP .B flakey -.br Creates a similar mapping to the linear target but exhibits unreliable behaviour periodically. Useful for simulating failing devices when testing. -.HP +.TP .B mirror -.br Mirrors data across two or more devices. -.HP +.TP .B multipath -.br Mediates access through multiple paths to the same device. -.HP +.TP .B raid -.br Offers an interface to the kernel's software raid driver, md. -.HP +.TP .B snapshot -.br Supports snapshots of devices. -.HP -.BR thin ,\ thin-pool -.br +.TP +.BR thin ", " thin-pool Supports thin provisioning of devices and also provides a better snapshot support. -.PD .P To find out more about the various targets and their table formats and status lines, please read the files in the Documentation/device-mapper directory in the kernel source tree. (Your distribution might include a copy of this information in the documentation directory for the device-mapper package.) - +. .SH EXAMPLES - +. # A table to join two disks together .br -.br 0 1028160 linear /dev/hda 0 .br 1028160 3903762 linear /dev/hdb 0 - - +.br # A table to stripe across the two disks, .br # and add the spare space from .br # hdb to the back of the volume - +.br 0 2056320 striped 2 32 /dev/hda 0 /dev/hdb 0 .br 2056320 2875602 linear /dev/hdb 1028160 - +. .SH ENVIRONMENT VARIABLES +. .TP .B DM_DEV_DIR The device directory name. -Defaults to "/dev" and must be an absolute path. +Defaults to "\fI/dev\fP" and must be an absolute path. .TP .B DM_UDEV_COOKIE A cookie to use for all relevant commands to synchronize with udev processing. @@ -787,15 +1008,17 @@ It is an alternative to using \fB\-\-udevcookie\fP option. .B DM_DEFAULT_NAME_MANGLING_MODE A default mangling mode. Defaults to "\fB#DEFAULT_MANGLING#\fP" and it is an alternative to using \fB\-\-manglename\fP option. - +. .SH AUTHORS -Original version: Joe Thornber (thornber(a)redhat.com) - +. +Original version: Joe Thornber <thornber(a)redhat.com> +. .SH SEE ALSO +. .BR dmstats (8), .BR udev (7), .BR udevadm (8) - -LVM2 resource page https://www.sourceware.org/lvm2/ +.P +LVM2 resource page: https://www.sourceware.org/lvm2/ .br Device-mapper resource page: http://sources.redhat.com/dm/ diff --git a/man/dmstats.8.in b/man/dmstats.8.in index 985f3fd..9776aa7 100644 --- a/man/dmstats.8.in +++ b/man/dmstats.8.in @@ -1,278 +1,420 @@ .TH DMSTATS 8 "Jul 25 2015" "Linux" "MAINTENANCE COMMANDS" + +.de OPT_PROGRAMS +. RB \%[ \-\-allprograms | \-\-programid +. IR id ] +.. +. +.de OPT_REGIONS +. RB \%[ \-\-allregions | \-\-regionid +. IR id ] +.. +. +.\" Print units suffix, use with arg to print human +.\" man2html can't handle too many changes per command +.de UNITS +. BR b | B | s | S | k | K | m | M | \c +. BR g | G | t | T | p | P | e | E ] +.. +. +.\" Print help text for units, use with arg to print human +.de HELP_UNITS +. RB ( b )ytes, +. RB ( s )ectors, +. RB ( k )ilobytes, +. RB ( m )egabytes, +. RB ( g )igabytes, +. RB ( t )erabytes, +. RB ( p )etabytes, +. RB ( e )xabytes. +. nop Capitalise to use multiples of 1000 (S.I.) instead of 1024. +.. +. .SH NAME +. dmstats \(em device-mapper statistics management +. .SH SYNOPSIS -.ad l -.B dmsetup stats +. +.B dmsetup +.B stats .I command .RB [ options ] -.HP +.sp +. .PD 0 +.HP .B dmstats -.I command -.RI [ device_name | \fB\-\-uuid -.IR uuid | \fB\-\-major -.I major -.B \-\-minor -.IR minor ] -.HP -.B dmstats clear -.I device_name -.RB [ \-\-allregions | \-\-regionid -.IR id ] -.HP -.B dmstats create -.I device_name -.RB [ \-\-alldevices ] -.RB [ \-\-areas -.IR nr_areas | \fB\-\-areasize -.IR area_size ] -.RB [ \-\-bounds -.IR histogram_boundaries ] -.RB [ \-\-precise ] -.RB [ \-\-start -.I start_sector -.B \-\-length -.IR length | \fB\-\-segments ] -.RB [ \-\-auxdata -.IR data ] -.RB [ \-\-programid -.IR id ] -.HP -.B dmstats delete -.I device_name -.RB [ \-\-alldevices ] -.RB [ \-\-allregions | \-\-regionid -.IR id ] -.RB [ \-\-allprograms | \-\-programid -.IR id ] -.HP -.B dmstats help -.RB [ \-c | \-C | \-\-columns ] -.HP -.B dmstats list -.RI [ device_name ] -.RB [ \-\-allprograms | \-\-programid -.IR id ] -.RB [ \-\-histogram ] -.RB [ \-\-units -.IR units ] -.RB [ \-\-nosuffix ] -.RB [ \-\-notimesuffix ] -.RB [ \-v | \-\-verbose \ [ \-v | \-\-verbose ]] -.HP -.B dmstats print -.RI [ device_name ] -.RB [ \-\-clear ] -.RB [ \-\-allprograms | \-\-programid -.IR id ] -.RB [ \-\-allregions | \-\-regionid -.IR id ] +.de CMD_COMMAND +. ad l +. IR command +. RI [ device_name +. RB | \-\-uuid +. IR uuid | \fB\-\-major +. IR major +. BR \-\-minor +. IR minor ] +. ad b +.. +.CMD_COMMAND +. .HP -.B dmstats report -.RI [ device_name ] -.RB [ \-\-interval -.IR seconds ] -.RB [ \-\-count -.IR count ] -.RB [ \-\-units -.IR units ] -.RB [ \-\-histogram ] -.RB [ \-\-allprograms ] -.RB [ \-\-programid -.IR id ] -.RB [ \-\-regionid -.IR id ] -.RB [ \-O | \-\-sort -.IR sort_fields ] -.RB [ \-S | \-\-select -.IR Selection ] -.RB [ \-\-units -.IR units ] -.RB [ \-\-nosuffix ] -.RB [ \-\-notimesuffix ] -.PD +.B dmstats +.de CMD_CLEAR +. ad l +. BR clear +. IR device_name +. OPT_PROGRAMS +. OPT_REGIONS +. ad b +.. +.CMD_CLEAR +. +.HP +.B dmstats +.de CMD_CREATE +. ad l +. BR create +. IR device_name +. RB [ \-\-alldevices ] +. RB [ \-\-areas +. IR nr_areas | \fB\-\-areasize +. IR area_size ] +. RB [ \-\-bounds +. IR \%histogram_boundaries ] +. RB [ \-\-precise ] +. RB [ \-\-start +. IR start_sector +. BR \-\-length +. IR length | \fB\-\-segments ] +. RB \%[ \-\-auxdata +. IR data ] +. RB [ \-\-programid +. IR id ] +. ad b +.. +.CMD_CREATE +. +.HP +.B dmstats +.de CMD_DELETE +. ad l +. BR delete +. RI [ device_name ] +. RB [ \-\-alldevices ] +. OPT_PROGRAMS +. OPT_REGIONS +. ad b +.. +.CMD_DELETE +. .HP +.B dmstats +.de CMD_HELP +. ad l +. BR help +. RB [ \-c | \-C | \-\-columns ] +. ad b +.. +.CMD_HELP +. +.HP +.B dmstats +.de CMD_LIST +. ad l +. BR list +. RI [ device_name ] +. RB [ \-\-histogram ] +. OPT_PROGRAMS +. RB [ \-\-units +. IR units ] +. RB \%[ \-\-nosuffix ] +. RB [ \-\-notimesuffix ] +. RB \%[ \-v | \-\-verbose [ \-v | \-\-verbose ]] +. ad b +.. +.CMD_LIST +. +.HP +.B dmstats +.de CMD_PRINT +. ad l +. BR print +. RI [ device_name ] +. RB [ \-\-clear ] +. OPT_PROGRAMS +. OPT_REGIONS +. ad b +.. +.CMD_PRINT +. +.HP +.B dmstats +.de CMD_REPORT +. ad l +. BR report +. RI [ device_name ] +. RB [ \-\-interval +. IR seconds ] +. RB [ \-\-count +. IR count ] +. RB [ \-\-units +. IR units ] +. RB [ \-\-histogram ] +. OPT_PROGRAMS +. OPT_REGIONS +. RB [ \-O | \-\-sort +. IR sort_fields ] +. RB [ \-S | \-\-select +. IR selection ] +. RB [ \-\-units +. IR units ] +. RB [ \-\-nosuffix ] +. RB \%[ \-\-notimesuffix ] +. ad b +.. +.CMD_REPORT +. +.PD .ad b +. .SH DESCRIPTION +. The dmstats program manages IO statistics regions for devices that use the device-mapper driver. Statistics regions may be created, deleted, listed and reported on using the tool. -The first argument to dmstats is a command. +The first argument to dmstats is a \fIcommand\fP. -The second argument is the device name, uuid, or major and minor -numbers. +The second argument is the \fIdevice name\fP, +\fIuuid\fP or \fImajor\fP and \fIminor\fP numbers. Further options permit the selection of regions, output format control, and reporting behaviour. -When the program is run using the 'dmstats' alias, the command +When the program is run using the '\fBdmstats\fP' alias, the command \fBmust\fP be the first argument and any switches and options should be specified following the command itself. This limitation is not present -when run as 'dmsetup stats'. +when run as '\fBdmsetup stats\fP'. When no device argument is given dmstats will by default operate on all device-mapper devices present. The \fBcreate\fP and \fBdelete\fP -commands require the use of \fB--alldevices\fP when used in this way. - +commands require the use of \fB\-\-alldevices\fP when used in this way. +. .SH OPTIONS -.TP -.B \-\-alldevices +. +.HP +.BR \-\-alldevices +.br If no device arguments are given allow operation on all devices when creating or deleting regions. -.TP -.B \-\-allprograms +. +.HP +.BR \-\-allprograms +.br Include regions from all program IDs for list and report operations. -.TP -.B \-\-allregions +.br +.HP +.BR \-\-allregions +.br Include all present regions for commands that normally accept a single region identifier. -.TP -.B \-\-areas \fInr_areas +. +.HP +.BR \-\-areas +.IR nr_areas +.br Specify the number of statistics areas to create within a new region. -.TP -.B \-\-areasize \fIarea_size +. +.HP +.BR \-\-areasize +.IR area_size \c +.RB [ \c +.UNITS +.br Specify the size of areas into which a new region should be divided. An -optional suffix selects units of bBsSkKmMgGtTpPeE: (b)ytes, -(s)ectors, (k)ilobytes, (m)egabytes, (g)igabytes, (t)erabytes, -(p)etabytes, (e)xabytes. Capitalise to use multiples of 1000 (S.I.) -instead of 1024. -.TP -.B \-\-auxdata \fIaux_data +optional suffix selects units of: +.HELP_UNITS +. +.HP +.BR \-\-auxdata +.IR aux_data +.br Specify auxilliary data (a string) to be stored with a new region. -.TP -.B \-\-clear +. +.HP +.BR \-\-clear +.br When printing statistics counters, also atomically reset them to zero. -.TP -.B \-\-count \fIcount +. +.HP +.BR \-\-count +.IR count +.br Specify the iteration count for repeating reports. If the count argument is zero reports will continue to repeat until interrupted. -.TP -.B \-\-bounds \fIhistogram_boundaries +. +.HP +.BR \-\-bounds +.IR histogram_boundaries \c +.RB [ ns | us | ms | s ] +.br Specify the boundaries of a latency histogram to be tracked for the region as a comma separated list of latency values. Latency values are -given in nanoseconds. An optional unit suffix of ns, us, ms, or s may be -given after each value to specify units of nanoseconds, microseconds, -miliseconds or seconds respectively. -.TP -.B \-\-histogram +given in nanoseconds. An optional unit suffix of +.BR ns , +.BR us , +.BR ms , +or \fBs\fP may be given after each value to specify units of +nanoseconds, microseconds, miliseconds or seconds respectively. +. +.HP +.BR \-\-histogram +.br When used with the \fBreport\fP and \fBlist\fP commands select default fields that emphasize latency histogram data. -.TP -.B \-\-interval \fIseconds +. +.HP +.BR \-\-interval +.IR seconds +.br Specify the interval in seconds between successive iterations for -repeating reports. If \-\-interval is specified but \-\-count is not, +repeating reports. If \fB\-\-interval\fP is specified but +\fB\-\-count\fP is not, reports will continue to repeat until interrupted. -.TP -.B \-\-length \fIlength +. +.HP +.BR \-\-length +.IR length \c +.RB [ \c +.UNITS +.br Specify the length of a new statistics region in sectors. An optional -suffix selects units of bBsSkKmMgGtTpPeE: (b)ytes, (s)ectors, -(k)ilobytes, (m)egabytes, (g)igabytes, (t)erabytes, (p)etabytes, -(e)xabytes. Capitalise to use multiples of 1000 (S.I.) instead of 1024. -.TP -.BR \-j | \-\-major\ \fImajor +suffix selects units of: +.HELP_UNITS +. +.HP +.BR \-j | \-\-major +.IR major +.br Specify the major number. -.TP -.BR \-m | \-\-minor\ \fIminor +. +.HP +.BR \-m | \-\-minor +.IR minor +.br Specify the minor number. -.TP -.B \-\-nosuffix +. +.HP +.BR \-\-nosuffix +.br Suppress the suffix on output sizes. Use with \fB\-\-units\fP (except h and H) if processing the output. -.TP -.B \-\-notimesuffix +. +.HP +.BR \-\-notimesuffix +.br Suppress the suffix on output time values. Histogram boundary values will be reported in units of nanoseconds. -.TP +. +.HP .BR \-o | \-\-options +.br Specify which report fields to display. -.TP -.BR \-O | \-\-sort\ \fIsort_fields +. +.HP +.BR \-O | \-\-sort +.IR sort_fields +.br Sort output according to the list of fields given. Precede any -sort_field with - for a reverse sort on that column. -.TP +sort field with '\fB-\fP' for a reverse sort on that column. +. +.HP .BR \-\-precise +.br Attempt to use nanosecond precision counters when creating new statistics regions. -.TP -.B \-\-programid \fIid +. +.HP +.BR \-\-programid +.IR id +.br Specify a program ID string. When creating new statistics regions this string is stored with the region. Subsequent operations may supply a program ID in order to select only regions with a matching value. The default program ID for dmstats-managed regions is "dmstats". -.TP -.B \-\-regionid \fIid +. +.HP +.BR \-\-regionid +.IR id +.br Specify the region to operate on. -.TP +. +.HP .BR \-\-relative +.br If displaying the histogram report show relative (percentage) values instead of absolute counts. -.TP -.BR \-S | \-\-select \ \fIselection -Display only rows that match selection criteria. All rows with the -additional "selected" column (-o selected) showing 1 if the row matches -the selection and 0 otherwise. The selection criteria are defined by +. +.HP +.BR \-S | \-\-select +.IR selection +.br +Display only rows that match \fIselection\fP criteria. All rows with the +additional "selected" column (\fB\-o selected\fP) showing 1 if the row matches +the \fIselection\fP and 0 otherwise. The selection criteria are defined by specifying column names and their valid values while making use of supported comparison operators. -.TP -.B \-\-start \fIstart +. +.HP +.BR \-\-start +.IR start \c +.RB [ \c +.UNITS +.br Specify the start offset of a new statistics region in sectors. An -optional suffix selects units of bBsSkKmMgGtTpPeE: (b)ytes, -(s)ectors, (k)ilobytes, (m)egabytes, (g)igabytes, (t)erabytes, -(p)etabytes, (e)xabytes. Capitalise to use multiples of 1000 (S.I.) -instead of 1024. -.TP -.B \-\-segments +optional suffix selects units of: +.HELP_UNITS +. +.HP +.BR \-\-segments +.br Create a new statistics region for each target contained in the target device. This causes a separate region to be allocated for each segment of the device. -.TP -.BR \-\-units \ hHbBsSkKmMgGtTpPeE -Set the display units for report output. All sizes are output in these -units: (h)uman-readable, (b)ytes, (s)ectors, (k)ilobytes, (m)egabytes, -(g)igabytes, (t)erabytes, (p)etabytes, (e)xabytes. Capitalise to use -multiples of 1000 (S.I.) instead of 1024. Can also specify custom units -e.g. \fB\-\-units 3M\fP -.TP +. +.HP +.BR \-\-units +.RI [ units ] \c +.RB [ h | H | \c +.UNITS +.br +Set the display units for report output. +All sizes are output in these units: +.RB ( h )uman-readable, +.HELP_UNITS +Can also specify custom units e.g. \fB\-\-units\ 3M\fP. +. +.HP .BR \-u | \-\-uuid +.br Specify the uuid. -.TP -.BR \-v | \-\-verbose \ [ \-v | \-\-verbose ] -Produce additional output. +. +.HP +.BR \-v | \-\-verbose " [" \-v | \-\-verbose ] .br +Produce additional output. +. .SH COMMANDS -.TP -.B clear -.I device_name -.RB [ \-\-allregions | \-\-regionid -.IR id ] -.RB [ \-\-allprograms -.RB | -.B \-\-programid -.IR id ] +. +.HP +.CMD_CLEAR .br Instructs the kernel to clear statistics counters for the speficied regions (with the exception of in-flight IO counters). -.br -.TP -.B create -.I device_name -.RB [ \-\-areas -.IR nr_areas ] -.RB [ \-\-areasize -.IR area_size ] -.RB [ --bounds -.IR histogram_boundaries ] -.RB [ \-\-precise ] -.RB [ \-\-start -.I start_sector -.B \-\-length -.IR length | \fB\-\-segments ] -.RB [ \-\-auxdata -.IR data ] -.RB [ \-\-programid -.IR id ] +. +.HP +.CMD_CREATE .br Creates one or more new statistics regions on the specified device(s). @@ -286,7 +428,7 @@ device-mapper device's table. If the \fB\-\-precise\fP option is used the command will attempt to create a region using nanosecond precision counters. -If \fB--bounds\fP is given a latency histogram will be tracked for +If \fB\-\-bounds\fP is given a latency histogram will be tracked for the new region. The boundaries of the histogram bins are given as a comma separated list of latency values. There is an implicit lower bound of zero on the first bin and an implicit upper bound of infinity (or the @@ -311,17 +453,9 @@ By default dmstats creates regions with a \fBprogram_id\fP of On success the \fBregion_id\fP of the newly created region is printed to stdout. -.br -.TP -.B delete -.RI [ device_name ] -.RB [ \-\-alldevices ] -.RB [ \-\-allregions -| -.B \-\-regionid -.IR id ] -.RB [ \-\-allprograms | \-\-programid -.IR id ] +. +.HP +.CMD_DELETE .br Delete the specified statistics region. All counters and resources used by the region are released and the region will not appear in the output @@ -330,24 +464,17 @@ of subsequent list, print, or report operations. All regions registered on a device may be removed using \fB\-\-allregions\fP. -To remove all regions on all devices both \fB--allregions\fP and +To remove all regions on all devices both \fB\-\-allregions\fP and \fB\-\-alldevices\fP must be used. -.br -.TP -.B help -.RB [ \-c | \-C | \-\-columns ] +. +.HP +.CMD_HELP .br Outputs a summary of the commands available, optionally including the list of report fields. -.br -.TP -.B list -.RI [ device_name ] -.RB [ \-\-histogram ] -.RB [ \-\-allprograms ] -.RB [ \-\-programid -.IR id ] -.RB [ \-v | \-\-verbose \ [ \-v | \-\-verbose ]] +. +.HP +.CMD_LIST .br List the statistics regions registered on the device. If the \fB\-\-allprograms\fP switch is given all regions will be listed @@ -358,37 +485,15 @@ a row of information for each area contained in each region displayed. If \fB\-\-histogram\fP is given the report will include the bin count and latency boundary values for any configured histograms. -.br -.TP -.B print -.RB [ \-\-allregions | \-\-regionid -.IR id ] -.RB [ \-\-allprograms | \-\-programid -.IR id ] -.RB [ \-\-clear ] +. +.HP +.CMD_PRINT .br Print raw statistics counters for the specified region or for all present regions. -.br -.TP -.B report -.RB [ \-\-allprograms ] -.RB [ \-\-interval -.IR seconds ] -.RB [ \-\-count -.IR count ] -.RB [ \-\-units -.IR unit ] -.RB [ \-\-regionid -.IR id ] -.RB [ \-\-programid -.IR id ] -.RB [ \-O | \-\-sort -.IR sort_fields ] -.RB [ \-S | \-\-select -.IR selection ] -.RB [ \-\-units -.IR units ] +. +.HP +.CMD_REPORT .br Start a report for the specified region or for all present regions. If the count argument is specified, the report will repeat at a fixed @@ -398,14 +503,14 @@ one second. If the \fB\-\-allprograms\fP switch is given, all regions will be listed, regardless of region program ID values. -If \fB\-\-histogram\fP is given the report will include the histogram +If the \fB\-\-histogram\fP is given the report will include the histogram values and latency boundaries. -If \fB\-\-relative\fP is used the default histogram field displays +If the \fB\-\-relative\fP is used the default histogram field displays bin values as a percentage of the total number of I/Os. - -.br +. .SH REGIONS AND AREAS +. The device-mapper statistics facility allows separate performance counters to be maintained for arbitrary regions of devices. A region may span any range: from a single sector to the whole device. A region may @@ -421,113 +526,96 @@ areas, or into areas of the given size by specifying one of \fB\-\-areas\fP or \fB\-\-areasize\fP when creating a region with the \fBcreate\fP command. Depending on the size of the areas and the device region the final area within the region may be smaller than requested. - -.SS Region identifiers +.P +.B Region identifiers +.P Each region is assigned an identifier when it is created that is used to reference the region in subsequent operations. Region identifiers are unique within a given device (including across different \fBprogram_id\fP values). -.br + Depending on the sequence of create and delete operations, gaps may exist in the sequence of \fBregion_id\fP values for a particular device. - +. .SH REPORT FIELDS +. The dmstats report provides several types of field that may be added to the default field set, or used to create custom reports. -.br + All performance counters and metrics are calculated per-area. -.br +. .SS Derived metrics +. A number of metrics fields are included that provide high level performance indicators. These are based on the fields provided by the conventional Linux iostat program and are derived from the basic counter values provided by the kernel for each area. -.br -.HP +.TP .B reads_merged_per_sec -.br Reads merged per second. -.HP +.TP .B writes_merged_per_sec -.br Writes merged per second. -.HP +.TP .B reads_per_sec -.br Reads completed per second. -.HP +.TP .B writes_per_sec -.br Writes completed per second. -.HP +.TP .B read_size_per_sec -.br Size of data read per second. -.HP +.TP .B write_size_per_sec -.br Size of data written per second. -.HP +.TP .B avg_request_size -.br Average request size. -.HP +.TP .B queue_size -.br Average queue size. -.HP +.TP .B await -.br The average wait time for read and write operations. -.HP +.TP .B r_await -.br The average wait time for read operations. -.HP +.TP .B w_await -.br The average wait time for write operations. -.HP +.TP .B throughput -.br The device throughput in operations per second. -.HP +.TP .B service_time -.br -The average service time (in milliseconds) for operations issued +The average service time (in milliseconds) for operations issued to the device. -.HP +.TP .B util -.br Percentage of CPU time during which I/O requests were issued to the device (bandwidth utilization for the device). Device saturation occurs when this value is close to 100%. -.br +. .SS Region and area meta fields +. Meta fields provide information about the region or area that the statistics values relate to. This includes the region and area identifier, start, length, and counts, as well as the program ID and auxiliary data values. -.br -.HP +.TP .B region_id -.br Region identifier. This is a non-negative integer returned by the kernel when a statistics region is created. -.HP +.TP .B region_start -.br -.br The region start location. Display units are selected by the -\fB--units\fP option. -.HP +\fB\-\-units\fP option. +.TP .B region_len -.br The length of the region. Display units are selected by the -\fB--units\fP option. -.HP +\fB\-\-units\fP option. +.TP .B area_id -.br Area identifier. Area identifiers are assigned by the device-mapper statistics library and uniquely identify each area within a region. Each ID corresponds to a distinct set of performance counters for that area @@ -535,206 +623,170 @@ of the statistics region. Area identifiers are always monotonically increasing within a region so that higher ID values correspond to greater sector addresses within the area and no gaps in the sequence of identifiers exist. -.HP +.TP .B area_start -.br The area start location. Display units are selected by the -\fB--units\fP option. -.HP +\fB\-\-units\fP option. +.TP .B area_len -.br The length of the area. Display units are selected by the -\fB--units\fP option. -.HP +\fB\-\-units\fP option. +.TP .B area_count -.br The number of areas in this region. -.HP +.TP .B program_id -.br The program ID value associated with this region. -.HP +.TP .B aux_data -.br The auxiliary data value associated with this region. -.br -.HP +.TP .B interval_ns -.br The estimated interval over which the current counter values have accumulated. The value is reported as an interger expressed in units of nanoseconds. -.br -.HP +.TP .B interval -.br The estimated interval over which the current counter values have accumulated. The value is reported as a real number in units of seconds. -.br +. .SS Basic counters +. Basic counters provide access to the raw counter data from the kernel, allowing further processing to be carried out by another program. - +.P The kernel provides thirteen separate counters for each statistics area. The first eleven of these match the counters provided in /proc/diskstats or /sys/block/*/*/stat. The final pair provide separate counters for read and write time. -.P -.HP +.TP .B read_count -.br Count of reads completed this interval. -.HP +.TP .B reads_merged_count -.br Count of reads merged this interval. -.HP +.TP .B read_sector_count -.br Count of 512 byte sectors read this interval. -.HP +.TP .B read_time -.br Accumulated duration of all read requests (ns). -.HP +.TP .B write_count -.br Count of writes completed this interval. -.HP +.TP .B writes_merged_count -.br Count of writes merged this interval. -.HP +.TP .B write_sector_count -.br Count of 512 byte sectors written this interval. -.HP +.TP .B write_nsecs -.br Accumulated duration of all write requests (ns). -.HP +.TP .B in_progress_count -.br Count of requests currently in progress. -.HP +.TP .B io_ticks -.br Nanoseconds spent servicing requests. -.HP +.TP .B queue_ticks -.br This field is incremented at each I/O start, I/O completion, I/O merge, or read of these stats by the number of I/Os in progress multiplied by the number of milliseconds spent doing I/O since the last update of this field. This can provide an easy measure of both I/O completion time and the backlog that may be accumulating. -.HP +.TP .B read_ticks -.br Nanoseconds spent servicing reads. -.HP +.TP .B write_ticks -.br Nanoseconds spent servicing writes. -.br +. .SS Histogram fields +. Histograms measure the frequency distribution of user specified I/O latency intervals. Histogram bin boundaries are specified when a region is created. - +.P A brief representation of the histogram values and latency intervals can be included in the report using these fields. -.P -.HP +.TP .B hist_count -.br A list of the histogram counts for the current statistics area in order of ascending latency value. Each value represents the number of I/Os with latency times falling into that bin's time range during the sample period. -.HP +.TP .B hist_count_bounds -.br A list of the histogram counts for the current statistics area in order of ascending latency value including bin boundaries: each count is prefixed by the lower bound of the corresponding histogram bin. -.HP +.TP .B hist_count_ranges -.br A list of the histogram counts for the current statistics area in order of ascending latency value including bin boundaries: each count is prefixed by both the lower and upper bounds of the corresponding histogram bin. -.HP +.TP .B hist_percent -.br A list of the relative histogram values for the current statistics area in order of ascending latency value, expressed as a percentage. Each value represents the proportion of I/Os with latency times falling into that bin's time range during the sample period. -.HP +.TP .B hist_percent_bounds -.br A list of the relative histogram values for the current statistics area in order of ascending latency value, expressed as a percentage and including bin boundaries. Each value represents the proportion of I/Os with latency times falling into that bin's time range during the sample period and is prefixed with the corresponding bin's lower bound. -.HP +.TP .B hist_percent_ranges -.br A list of the relative histogram values for the current statistics area in order of ascending latency value, expressed as a percentage and including bin boundaries. Each value represents the proportion of I/Os with latency times falling into that bin's time range during the sample period and is prefixed with the corresponding bin's lower and upper bounds. -.HP +.TP .B hist_bounds -.br A list of the histogram boundary values for the current statistics area in order of ascending latency value. The values are expressed in whole units of seconds, miliseconds, microseconds or nanoseconds with a suffix indicating the unit. -.HP +.TP .B hist_ranges -.br A list of the histogram bin ranges for the current statistics area in order of ascending latency value. The values are expressed as "LOWER-UPPER" in whole units of seconds, miliseconds, microseconds or nanoseconds with a suffix indicating the unit. -.HP +.TP .B hist_bins -.br The number of latency histogram bins configured for the area. -.br -.br -.P +. .SH EXAMPLES +. Create a whole-device region with one area on vg00/lvol1 .br -.br -# dmstats create vg00/lvol1 +# +.B dmstats create vg00/lvol1 .br vg00/lvol1: Created new region with 1 area(s) as region ID 0 -.br -.br - - +.P Create a 32M region 1G into device d0 .br -.br -# dmstats create --start 1G --length 32M d0 +# +.B dmstats create \-\-start 1G \-\-length 32M d0 .br d0: Created new region with 1 area(s) as region ID 0 -.br - - +.P Create a whole-device region with 8 areas on every device .br .br -# dmstats create --areas 8 +# +.B dmstats create \-\-areas 8 .br vg00-lvol1: Created new region with 8 area(s) as region ID 0 .br @@ -747,37 +799,33 @@ vg01-lvol0: Created new region with 8 area(s) as region ID 2 vg01-lvol1: Created new region with 8 area(s) as region ID 0 .br vg00-lvol2: Created new region with 8 area(s) as region ID 1 -.br -.br - +.P Delete all regions on all devices .br .br -# dmstats delete --alldevices --allregions -.br -.br - +# +.B dmstats delete \-\-alldevices \-\-allregions +.P Create a whole-device region with areas 10GiB in size on vg00/lvol1 using dmsetup .br .br -# dmsetup stats create --areasize 10G vg00/lvol1 +# +.B dmsetup stats create \-\-areasize 10G vg00/lvol1 .br vg00-lvol1: Created new region with 5 area(s) as region ID 1 -.br -.br - +.P Create a 1GiB region with 16 areas at the start of vg00/lvol1 .br -# dmstats create --start 0 --len 1G --areas=16 vg00/lvol1 +# +.B dmstats create \-\-start 0 \-\-len 1G \-\-areas=16 vg00/lvol1 .br vg00-lvol1: Created new region with 16 area(s) as region ID 0 -.br -.br - +.P List the statistics regions registered on vg00/lvol1 .br -# dmstats list vg00/lvol1 +# +.B dmstats list vg00/lvol1 .br Name RgID RStart RSize #Areas ASize ProgID .br @@ -786,15 +834,15 @@ vg00-lvol1 0 0 61.00g 1 61.00g dmstats vg00-lvol1 1 61.00g 19.20g 1 19.20g dmstats .br vg00-lvol1 2 80.20g 2.14g 1 2.14g dmstats -.br -.br - +.P Display five statistics reports for vg00/lvol1 at an interval of one second .br .br -# dmstats report --interval 1 --count 5 vg00/lvol1 +# +.B dmstats report \-\-interval 1 \-\-count 5 vg00/lvol1 .br -# dmstats report +# +.B dmstats report .br Name RgID ArID AStart ASize RRqM/s WRqM/s R/s W/s RSz/s WSz/s AvRqSz QSize Util% AWait RdAWa WrAWa .br @@ -803,41 +851,39 @@ vg_hex-lv_home 0 0 0 61.00g 0.00 0.00 0.00 218.00 0 vg_hex-lv_home 1 0 61.00g 19.20g 0.00 0.00 0.00 5.00 0 548.00k 109.50k 0.14 11.00 27.40 0.00 27.40 .br vg_hex-lv_home 2 0 80.20g 2.14g 0.00 0.00 0.00 14.00 0 1.15m 84.00k 0.39 18.70 27.71 0.00 27.71 -.br -.br - +.P Create one region for reach target contained in device vg00/lvol1 .br .br -# dmstats create --segments vg00/lvol1 +# +.B dmstats create \-\-segments vg00/lvol1 .br vg00-lvol1: Created new region with 1 area(s) as region ID 0 .br vg00-lvol1: Created new region with 1 area(s) as region ID 1 .br vg00-lvol1: Created new region with 1 area(s) as region ID 2 -.br -.br - +.P Print raw counters for region 4 on device d0 .br -.br -# dmstats print --regionid 4 d0 +# +.B dmstats print \-\-regionid 4 d0 .br 2097152+65536 0 0 0 0 29 0 264 701 0 41 701 0 41 -.br -.br +. .SH AUTHORS +. Bryn M. Reeves <bmr(a)redhat.com> - +. .SH SEE ALSO +. .BR dmsetup (8) -LVM2 resource page https://www.sourceware.org/lvm2/ +LVM2 resource page: https://www.sourceware.org/lvm2/ .br Device-mapper resource page: http://sources.redhat.com/dm/ .br Device-mapper statistics kernel documentation .br -Documentation/device-mapper/statistics.txt +.I Documentation/device-mapper/statistics.txt

1 0

master - alloc: fix update or area_len
by Zdenek Kabelac 01 Oct '15

01 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=a139275eca4fadca… Commit: a139275eca4fadca4a456fd27e8b2ba30d7d2c02 Parent: efcb3bbc8d2ebcc39031cc0d7b446daf864d281e Author: Zdenek Kabelac <zkabelac(a)redhat.com> AuthorDate: Thu Oct 1 14:59:21 2015 +0200 Committer: Zdenek Kabelac <zkabelac(a)redhat.com> CommitterDate: Thu Oct 1 15:02:49 2015 +0200 alloc: fix update or area_len Commit: 192d9ad97769f2c16f3ebe67e51cfb287e67f67c changed logic for area_len formula - so it returns different values. Placing () to restore previous behaviour and make it explicit. --- lib/metadata/lv_manip.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/lib/metadata/lv_manip.c b/lib/metadata/lv_manip.c index 02a43b7..48d7d34 100644 --- a/lib/metadata/lv_manip.c +++ b/lib/metadata/lv_manip.c @@ -1932,7 +1932,7 @@ static int _for_each_pv(struct cmd_context *cmd, struct logical_volume *lv, *max_seg_len = remaining_seg_len; area_multiple = _calc_area_multiple(seg->segtype, seg->area_count, 0); - area_len = remaining_seg_len / (area_multiple ? : 1); + area_len = (remaining_seg_len / area_multiple) ? : 1; /* For striped mirrors, all the areas are counted, through the mirror layer */ if (top_level_area_index == -1)

1 0

master - test: Fix timeout handling
by Marian Csontos 01 Oct '15

01 Oct '15

Gitweb: http://git.fedorahosted.org/git/?p=lvm2.git;a=commitdiff;h=efcb3bbc8d2ebcc3… Commit: efcb3bbc8d2ebcc39031cc0d7b446daf864d281e Parent: 309979d57891912b977edbd89970bf25614ee9eb Author: Marian Csontos <mcsontos(a)redhat.com> AuthorDate: Thu Oct 1 13:17:00 2015 +0200 Committer: Marian Csontos <mcsontos(a)redhat.com> CommitterDate: Thu Oct 1 13:19:29 2015 +0200 test: Fix timeout handling Timeouts were considered as warnings only. Signalling failure is preferred. --- test/lib/brick-shelltest.h | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/test/lib/brick-shelltest.h b/test/lib/brick-shelltest.h index b29e626..0659d8e 100644 --- a/test/lib/brick-shelltest.h +++ b/test/lib/brick-shelltest.h @@ -267,7 +267,7 @@ struct Journal { std::cout << std::endl << "### " << status.size() << " tests: " << count( PASSED ) << " passed, " << count( SKIPPED ) << " skipped, " - << count( TIMEOUT ) + count( WARNED ) << " broken, " + << count( TIMEOUT ) << " timed out, " << count( WARNED ) << " warned, " << count( FAILED ) << " failed" << std::endl; } @@ -1047,7 +1047,7 @@ struct Main { if ( die || fatal_signal ) return 1; - return journal.count( Journal::FAILED ) ? 1 : 0; + return journal.count( Journal::FAILED ) || journal.count( Journal::TIMEOUT ) ? 1 : 0; } Main( Options o ) : die( false ), journal( o.outdir ), options( o ) {}

1 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

lvm2-commits October 2015