Wrong default multipath configuration for EL6
by Patrick Hurrelmann
Hi all,
after suffering from boot-failures after kernel-updates on my EL6 hosts
(the ones that san boot via iscsi), I finally managed to locate the root
cause of the problem: the default values as set by vdsm are incorrect
for EL6 and prohibit a successful iscsi boot (the faulty multipath.conf
gets embedded in the new initrd on a kernel update). Additionally the
following errors are repeatedly logged on all my hosts (device ids may
vary) with the default vdsm multipath config:
device-mapper: table: 253:5: multipath: error getting device
device-mapper: ioctl: error adding target to table
device-mapper: table: 253:5: multipath: error getting device
device-mapper: ioctl: error adding target to table
And it all boils down to the default value set for option getuid_callout
in multipath.conf. Vdsm sets it to the following:
getuid_callout "/sbin/scsi_id --whitelisted --replace-whitespace
--device=/dev/%n"
According to the man-page and other docu the default value on EL6 is
"/lib/udev/scsi_id --whitelisted --device=/dev/%n". Although the binary
/sbin/scsi_id is a valid link to the target /lib/udev/scsi_id, the link
itself (/sbin/scsi_id) is _not_ included in the generated initrd. The
binary /lib/udev/scsi_id is indeed included and changing the default
config to use /lib/udev/scsi_id instead does make it all work again.
iscsi boot (after regenerating the initrd, as the multipath.conf is
embedded) is back to good and the previously logged device-mapper errors
are gone, too.
Can someone please verify this and and apply the easy (attached) patch?
--- multipath.py.orig 2014-07-29 15:24:37.643839169 +0200
+++ multipath.py 2014-07-29 15:25:06.823840362 +0200
@@ -88,9 +88,8 @@
log = logging.getLogger("Storage.Multipath")
_scsi_id = utils.CommandPath("scsi_id",
- "/sbin/scsi_id", # EL6
"/usr/lib/udev/scsi_id", # Fedora
- "/lib/udev/scsi_id", # Ubuntu
+ "/lib/udev/scsi_id", # EL6, Ubuntu
)
Thanks a lot and best regards
Patrick
--
Lobster SCM GmbH, Münchner Straße 15a, D-82319 Starnberg
HRB 178831, Amtsgericht München
Geschäftsführer: Dr. Martin Fischer, Rolf Henrich
9 years, 8 months
Bugzilla bugs moved to MODIFIED
by dcaroest＠redhat.com
Hi!
Yesterday I made some changes to the gerrit hooks to avoid having bugs
moved to MODIFIED status when the patch that was merged was in a
master branch.
That should not be happening again, if you see it happening, ping me
with the bug and the patch id and I'll look into it and solve it.
Thanks!
--
David Caro
Red Hat S.L.
Continuous Integration Engineer - EMEA ENG Virtualization R&D
Tel.: +420 532 294 605
Email: dcaro(a)redhat.com
Web: www.redhat.com
RHT Global #: 82-62605
9 years, 8 months
compile vdsm and attach it to a engine.
by aaron Beein
Hi,
Thank you for your great job on ovirt and vdsm. Now I devote myself to
compile vdsm on centos 6.3 host and attach it to a ovirt engine. But when I
attach the host which contains a compiled vdsm to a ovirt engine , the
status of the host is always ‘Non Responsive’(step 11 below). I reference
the links below:
http://www.ovirt.org/Vdsm_Developers
http://www.ovirt.org/Installing_VDSM_from_rpm
The steps( 1-9 ) are executed on centos6.3 host, and the steps(10--11) are
executed on ovirt engine. So I would be very grateful if you can give me
some clues that if I've missed anything or I done something wrong.
The attachment is the same as the bellow which makes it easier for you to
read.
1 Deployment platform
Centos6.3
Linux bogon 2.6.32-431.20.3.el6.x86_64 #1 SMP Thu Jun 19 21:14:45 UTC 2014
x86_64 x86_64 x86_64 GNU/Linux
Ip ： 10.1.8.252
CPU supports hardware virtualization extensions:
# cat /proc/cpuinfo | egrep 'svm|vmx'| grep nx
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca
cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx
rdtscp lm constant_tsc arch_perfmon pebs bts rep_good xtopology nonstop_tsc
aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr
pdcm pcid sse4_1 sse4_2 popcnt tsc_deadline_timer xsave lahf_lm arat epb
xsaveopt pln pts dts tpr_shadow vnmi flexpriority ept vpid
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca
cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx
rdtscp lm constant_tsc arch_perfmon pebs bts rep_good xtopology nonstop_tsc
aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr
pdcm pcid sse4_1 sse4_2 popcnt tsc_deadline_timer xsave lahf_lm arat epb
xsaveopt pln pts dts tpr_shadow vnmi flexpriority ept vpid
2 Apply all updates # yum -y update 3 Installing required packages
RHEL 6 users must add EPEL yum repository for installing python-ordereddict
and pyton-pthreading. The rpm bellow will install the epel yum repo and
required gpg keys.
# yum install http://download.fedoraproject.org/pub/epel/6/i386/epel-release-6-8.noarch...
RHEL 6 users must install a newer pep8 version than the one shipped in
EPEL6. Older pep8 versions have a bug that's tickled by vdsm. You can use
`pip`, or
yum install http://danken.fedorapeople.org/python-pep8-1.4.5-2.el6.noarch.rpm
oVirt repo:
yum install http://resources.ovirt.org/releases/ovirt-release.noarch.rpm
RHEL 6 users must add the glusterfs repository, providing newer glusterfs
not available on RHEL 6. Optionally install 'wget' if not present
rpm -q wget 2> /dev/null || yum install wget
wget -O /etc/yum.repos.d/glusterfs-epel.repo
*http://download.gluster.org/pub/gluster/glusterfs/LATEST/EPEL.repo/glusterfs-epel.repo*
<http://download.gluster.org/pub/gluster/glusterfs/LATEST/EPEL.repo/gluste...>
Fedora and RHEL 6 users must verify the following packages are installed
before attempting to build:
yum install make autoconf automake pyflakes logrotate gcc python-pep8
libvirt-python python-devel \
python-nose rpm-build sanlock-python genisoimage python-ordereddict
python-pthreading libselinux-python\
python-ethtool m2crypto python-dmidecode python-netaddr
python-inotify python-argparse git \
python-cpopen bridge-utils libguestfs-tools-c pyparted openssl libnl
libtool gettext-devel python-ioprocess libvirt libvirt-client
libvirt-lock-sanlock
4 Getting the source
cd /root
git clone *http://gerrit.ovirt.org/p/vdsm.git*
<http://gerrit.ovirt.org/p/vdsm.git>
cd vdsm
5 Building a Vdsm RPM
./autogen.sh –system
./configure --prefix=/usr --sysconfdir=/etc --localstatedir=/var
--libdir=/usr/lib --enable-hooks
make rpm NOSE_EXCLUDE=.*
6 Basic installation and start
When building from source, you should enable the ovirt-beta repository, to
satisfy dependencies that are not available yet in the release repository.
# cd ~/rpmbuild/RPMS
# yum install --skip-broken
--enablerepo=ovirt-master-snapshot-static x86_64/* noarch/vdsm-xml*
noarch/vdsm-cli* noarch/vdsm-python-zombiereaper*
noarch/vdsm-*jsonrpc*
Before starting vdsmd service for the first time vdsm requires some
configuration procedures for external services that being used by vdsmd. To
ease this process vdsm provides a utility (vdsm-tool). To perform full
reconfiguration of external services perform:
# vdsm-tool configure --force
(for more information read "vdsm-tool --help")
7 Finally start the vdsmd service
# service vdsmd start
8 Yum install -y bridge-utils
Configuring the bridge Interface as below
Disable the network manager service by executing as root:
systemctl stop NetworkManager.service
systemctl disable NetworkManager.service
service network start
chkconfig network on
Add the following content into a new file named:
*/etc/sysconfig/network-scripts/ifcfg-ovirtmgmt*:
DEVICE=ovirtmgmt
TYPE=Bridge
ONBOOT=yes
DELAY=0
BOOTPROTO=static
IPADDR=10.1.8.252
NETMASK=255.255.255.0
GATEWAY=10.1.8.254
Add the following line into the configuration file of your out going
interface (usually em1/eth0) the file is located at:
*/etc/sysconfig/network-scripts/ifcfg-em1* (assuming the device is em1)
BRIDGE=ovirtmgmt
and remove the IPADDR, NETMASK and BOOTPROTO keys, since the interface
should not have an IP address of its own. Full Example
DEVICE=em1
ONBOOT=yes
BRIDGE=ovirtmgmt
Restart the network service by executing:
service network restart
*Note that if any other bridge (from ovirtmgmt) is present at the time of
host installation, the bridge creation operation is skipped and you have to
change the bridge settings to correspond to above shown configuration
manually.*
*9 Configuring VDSM*
Add the following content into the file: */etc/vdsm/vdsm.conf* (you may
need to create that file):
[vars]
ssl = false
Restart the vdsmd service by executing:
service vdsmd restart
If Vdsm was started earlier with ssl=true, it would refuse to start and you
may need to use the undocumented verb
service vdsmd reconfigure
service vdsmd start
which edits */etc/libvirt/qemu.conf* and changes *spice_tls=1* to
*spice_tls=0*.
10 Connect to overt-engine
*ref: OVirt_-_connecting_development_vdsm_to_ovirt_engine
<http://www.ovirt.org/OVirt_-_connecting_development_vdsm_to_ovirt_engine>.*
*su - postgres -c "psql engine -c \"UPDATE vdc_options set option_value =
'true' where option_name = 'InstallVds'\""*
*service overt-engine restart*
11 Attach the host to the engine
I login the engine Administration Portal , and attach the centos host to a
cluster. But it failed.
9 years, 9 months
