anaconda-patches December 2015

anaconda-patches@lists.fedorahosted.org

9 participants
498 discussions

Start a nNew thread

Re: [rhinstaller/lorax/pulls/78 master] Add --image-type and --qemu-args
by dashea 18 Dec '15

18 Dec '15

Added label: master. -- To view this pull request on github, visit https://github.com/rhinstaller/lorax/pull/78

1 0

[master 3/3] docs: Update documentation for livemedia-creator
by dashea 18 Dec '15

18 Dec '15

From: "Brian C. Lane" <bcl(a)redhat.com> --- docs/livemedia-creator.1 | 8 ++++---- docs/livemedia-creator.rst | 8 ++++---- 2 files changed, 8 insertions(+), 8 deletions(-) diff --git a/docs/livemedia-creator.1 b/docs/livemedia-creator.1 index c6d2056..e1b5fc2 100644 --- a/docs/livemedia-creator.1 +++ b/docs/livemedia-creator.1 @@ -9,7 +9,7 @@ livemedia-creator [-h] [--fs-image FS_IMAGE] [--ks KS] [--image-name IMAGE_NAME] [--image-only] [--fs-label FS_LABEL] - [--qcow2] [--qcow2-arg QCOW2_ARGS] + [--image-type IMAGE_TYPE] [--qemu-arg QEMU_ARGS] [--compression] [--compress-arg] [--live-rootfs-size LIVE_ROOTFS_SIZE] [--live-rootfs-keep-size LIVE_ROOTFS_KEEP_SIZE] @@ -94,11 +94,11 @@ Path to disk image to use for creating final image Path to existing filesystem image to use for creating final image. .TP -\fB\-\-qcow2\fR -Create qcow2 image instead of raw sparse image when making disk images. +\fB\-\-image-type IMAGE_TYPE\fR +Create a qemu-img image instead of raw sparse image when making disk images. See qemu-img --help for a list of supported types. .TP -\fB\-\-qcow2\-arg\fR +\fB\-\-qemu\-arg\fR Arguments to pass to qemu-img. Pass once for each argument .TP diff --git a/docs/livemedia-creator.rst b/docs/livemedia-creator.rst index f739e64..3097e88 100644 --- a/docs/livemedia-creator.rst +++ b/docs/livemedia-creator.rst @@ -287,10 +287,10 @@ The created image can be imported into libvirt using: ``virt-image appliance.xml`` -You can also create qcow2 appliance images using ``--qcow2``, for example:: +You can also create qcow2 appliance images using ``--image-type=qcow2``, for example:: sudo livemedia-creator --make-appliance --iso=/path/to/boot.iso --ks=./docs/fedora-minimal.ks \ - --qcow2 --app-file=minimal-test.xml --image-name=minimal-test.img + --image-type=qcow2 --app-file=minimal-test.xml --image-name=minimal-test.img Filesystem Image Creation @@ -414,10 +414,10 @@ cloud-utils-growpart will grow the image to fit the instance's disk size. Create a qcow2 image using the kickstart like this: - ``sudo livemedia-creator --make-disk --iso=/path/to/boot.iso --ks=/path/to/fedora-openstack.ks --qcow2`` + ``sudo livemedia-creator --make-disk --iso=/path/to/boot.iso --ks=/path/to/fedora-openstack.ks --image-type=qcow2`` .. note:: - On the RHEL7 version of lmc ``--qcow2`` isn't supported. You can only create a bare partitioned disk image. + On the RHEL7 version of lmc ``--image-type`` isn't supported. You can only create a bare partitioned disk image. Import the resulting disk image into the OpenStack system, either via the web UI, or glance on the cmdline:: -- To view this commit on github, visit https://github.com/rhinstaller/lorax/commit/8e9dfcedce9bd8a393d793e0a1e3860…

1 0

[master 2/3] livemedia-creator: Add --image-type and --qemu-args options
by dashea 18 Dec '15

18 Dec '15

From: "Brian C. Lane" <bcl(a)redhat.com> Also alias --qcow2 to --image-type=qcow2 This allows --make-disk to be used to create any disk image that qemu-img supports, not just raw or qcow2. See qemu-img --help for a list of the supported image types. --- src/sbin/livemedia-creator | 84 ++++++++++++++++++++++++++-------------------- 1 file changed, 47 insertions(+), 37 deletions(-) diff --git a/src/sbin/livemedia-creator b/src/sbin/livemedia-creator index 98b76c7..a5af81c 100755 --- a/src/sbin/livemedia-creator +++ b/src/sbin/livemedia-creator @@ -52,7 +52,7 @@ from pylorax.treebuilder import findkernels from pylorax.sysutils import joinpaths, remove from pylorax.imgutils import PartitionMount, mksparse, mkext4img, loop_detach from pylorax.imgutils import get_loop_name, dm_detach, mount, umount, Mount -from pylorax.imgutils import mksquashfs, mkqcow2, mktar, mkrootfsimg +from pylorax.imgutils import mksquashfs, mkqemu_img, mktar, mkrootfsimg from pylorax.imgutils import copytree from pylorax.executils import execWithRedirect, execReadlines, runcmd from pylorax.monitor import LogMonitor @@ -99,7 +99,7 @@ class VirtualInstall(object): def __init__(self, iso, ks_paths, disk_img, img_size=2048, kernel_args=None, memory=1024, vcpus=None, vnc=None, arch=None, log_check=None, virtio_host="127.0.0.1", virtio_port=6080, - qcow2=False, boot_uefi=False, ovmf_path=None): + image_type=None, boot_uefi=False, ovmf_path=None): """ Start the installation @@ -118,7 +118,7 @@ class VirtualInstall(object): :type log_check: method :param str virtio_host: Hostname to connect virtio log to :param int virtio_port: Port to connect virtio log to - :param bool qcow2: Set to True if disk_img is a qcow2 + :param str image_type: Type of qemu-img disk to create, or None. :param bool boot_uefi: Use OVMF to boot the VM in UEFI mode :param str ovmf_path: Path to the OVMF firmware """ @@ -142,8 +142,8 @@ class VirtualInstall(object): disk_opts = "path={0}".format(disk_img) disk_opts += ",cache=unsafe,discard=unmap" - if qcow2: - disk_opts += ",format=qcow2" + if image_type: + disk_opts += ",format={0}".format(image_type) else: disk_opts += ",format=raw" if not os.path.isfile(disk_img): @@ -611,7 +611,7 @@ def novirt_install(opts, disk_img, disk_size, repo_url): :param str repo_url: The url of the repository to use for the installation This method makes sure SELinux is permissive during the install, runs anaconda - to create the image and then based on the opts passed create a qcow2 image + to create the image and then based on the opts passed create a qemu disk image or tarfile. """ import selinux @@ -713,20 +713,20 @@ def novirt_install(opts, disk_img, disk_size, repo_url): if selinux_enforcing: selinux.security_setenforce(1) - # qcow2 is used by bare qcow2 images and by Vagrant - if opts.qcow2: - log.info("Converting %s to qcow2", disk_img) - qcow2_args = [] - for arg in opts.qcow2_args: - qcow2_args += arg.split(" ", 1) - - # convert the image to qcow2 format - if "-O" not in qcow2_args: - qcow2_args.extend(["-O", "qcow2"]) - qcow2_img = tempfile.mktemp(prefix="disk", suffix=".img") - execWithRedirect("qemu-img", ["convert"] + qcow2_args + [disk_img, qcow2_img], raise_err=True) + # qemu disk image is used by bare qcow2 images and by Vagrant + if opts.image_type: + log.info("Converting %s to %s", disk_img, opts.image_type) + qemu_args = [] + for arg in opts.qemu_args: + qemu_args += arg.split(" ", 1) + + # convert the image to the selected format + if "-O" not in qemu_args: + qemu_args.extend(["-O", opts.image_type]) + qemu_img = tempfile.mktemp(prefix="disk", suffix=".img") + execWithRedirect("qemu-img", ["convert"] + qemu_args + [disk_img, qemu_img], raise_err=True) if not opts.make_vagrant: - execWithRedirect("mv", ["-f", qcow2_img, disk_img], raise_err=True) + execWithRedirect("mv", ["-f", qemu_img, disk_img], raise_err=True) else: # Take the new qcow2 image and package it up for Vagrant compress_args = [] @@ -735,7 +735,7 @@ def novirt_install(opts, disk_img, disk_size, repo_url): vagrant_dir = tempfile.mkdtemp() metadata_path = joinpaths(vagrant_dir, "metadata.json") - execWithRedirect("mv", ["-f", qcow2_img, joinpaths(vagrant_dir, "box.img")], raise_err=True) + execWithRedirect("mv", ["-f", qemu_img, joinpaths(vagrant_dir, "box.img")], raise_err=True) if opts.vagrant_metadata: shutil.copy2(opts.vagrant_metadata, metadata_path) else: @@ -796,13 +796,15 @@ def virt_install(opts, install_log, disk_img, disk_size): if opts.proxy: kernel_args += " proxy="+opts.proxy - if opts.qcow2 and not opts.make_fsimage: + if opts.image_type and not opts.make_fsimage: # virt-install can't take all the qcow2 options so create the image first - qcow2_args = [] - for arg in opts.qcow2_args: - qcow2_args += arg.split(" ", 1) + qemu_args = [] + for arg in opts.qemu_args: + qemu_args += arg.split(" ", 1) + if "-f" not in qemu_args: + qemu_args += ["-f", opts.image_type] - mkqcow2(disk_img, disk_size*1024**2, qcow2_args) + mkqemu_img(disk_img, disk_size*1024**2, qemu_args) if opts.make_fsimage or opts.make_tar or opts.make_oci: diskimg_path = tempfile.mktemp(prefix="disk", suffix=".img") @@ -815,7 +817,7 @@ def virt_install(opts, install_log, disk_img, disk_size): log_check = log_monitor.server.log_check, virtio_host = log_monitor.host, virtio_port = log_monitor.port, - qcow2=opts.qcow2, boot_uefi=opts.virt_uefi, + image_type=opts.image_type, boot_uefi=opts.virt_uefi, ovmf_path=opts.ovmf_path) virt.destroy() @@ -1098,10 +1100,14 @@ def main(): help="Name of output file to create. Used for tar, fs and disk image. Default is a random name.") image_group.add_argument("--fs-label", default="Anaconda", help="Label to set on fsimage, default is 'Anaconda'") + image_group.add_argument("--image-type", default=None, + help="Create an image with qemu-img. See qemu-img --help for supported formats.") + image_group.add_argument("--qemu-arg", action="append", dest="qemu_args", default=[], + help="Arguments to pass to qemu-img. Pass once for each argument, they will be used for ALL calls to qemu-img.") image_group.add_argument("--qcow2", action="store_true", help="Create qcow2 image instead of raw sparse image when making disk images.") - image_group.add_argument("--qcow2-arg", action="append", dest="qcow2_args", default=[], - help="Arguments to pass to qemu-img. Pass once for each argument") + image_group.add_argument("--qcow2-arg", action="append", dest="qemu_args", default=[], + help="Arguments to pass to qemu-img. Pass once for each argument, they will be used for ALL calls to qemu-img.") image_group.add_argument("--compression", default="xz", help="Compression binary for make-tar. xz, lzma, gzip, and bzip2 are supported. xz is the default.") image_group.add_argument("--compress-arg", action="append", dest="compress_args", default=[], @@ -1238,19 +1244,23 @@ def main(): # Vagrant creates a qcow2 inside a tar, turn on qcow2 if opts.make_vagrant: - opts.qcow2 = True + opts.image_type = "qcow2" + + # Alias --qcow2 to --image-type=qcow2 + if opts.qcow2: + opts.image_type = "qcow2" - if opts.qcow2 and not os.path.exists("/usr/bin/qemu-img"): - errors.append("qcow2 requires the qemu-img utility to be installed.") + if opts.image_type and not os.path.exists("/usr/bin/qemu-img"): + errors.append("image-type requires the qemu-img utility to be installed." % opts.image_type) - if opts.qcow2 and opts.make_iso: - errors.append("qcow2 cannot be used to make a bootable iso.") + if opts.image_type and opts.make_iso: + errors.append("image-type cannot be used to make a bootable iso.") - if opts.make_fsimage and opts.qcow2: - errors.append("qcow2 cannot be used to make filesystem images.") + if opts.image_type and opts.make_fsimage: + errors.append("image-type cannot be used to make filesystem images.") - if opts.make_tar and opts.qcow2: - errors.append("qcow2 cannot be used to make a tar.") + if opts.image_type and opts.make_tar: + errors.append("image-type cannot be used to make a tar.") if opts.make_oci and not (opts.oci_config and opts.oci_runtime): errors.append("--make-oci requires --oci-config and --oci-runtime") -- To view this commit on github, visit https://github.com/rhinstaller/lorax/commit/7705ff9ad5cf38aeca502591fe0c289…

1 0

[master 1/3] pylorax: Add mkqemu_img function, alias mkqcow2 to it.
by dashea 18 Dec '15

18 Dec '15

From: "Brian C. Lane" <bcl(a)redhat.com> Preserve the API but add a more descriptive name for the function. --- src/pylorax/imgutils.py | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/src/pylorax/imgutils.py b/src/pylorax/imgutils.py index 114e0de..eaaf1a3 100644 --- a/src/pylorax/imgutils.py +++ b/src/pylorax/imgutils.py @@ -133,6 +133,15 @@ def mkqcow2(outfile, size, options=None): Default format is qcow2, override by passing "-f", fmt in options. ''' + mkqemu_img(outfile, size, options) + +def mkqemu_img(outfile, size, options=None): + '''use qemu-img to create a file of the given size. + options is a list of options passed to qemu-img + + Default format is qcow2, override by passing "-f", fmt + in options. + ''' options = options or [] if "-f" not in options: options.extend(["-f", "qcow2"]) -- To view this commit on github, visit https://github.com/rhinstaller/lorax/commit/6e1c5af6d2f68712dc2352f5d609fda…

1 0

New: [rhinstaller/lorax/pulls/78 master] Add --image-type and --qemu-args
by dashea 18 Dec '15

18 Dec '15

This keeps --qcow2, and adds --image-type to allow other disk image types to be created. The old --qcow2 is aliased to --image-type=qcow2 and --qemu-arg is the same as --qcow2-arg -- To view this pull request on github, visit https://github.com/rhinstaller/lorax/pull/78

1 0

[rhel6-branch 1/1] Fix multipath UI representation (#914637)
by dashea 18 Dec '15

18 Dec '15

From: Robert Marshall <rmarshall(a)redhat.com> Changed the depopulate method of the FilterWindow class to correctly identify and remove outdated representations of a multipath device in the user interface. Added a check to ensure the first member device of a multipath device is not accidentally displayed and counted as both an mpath device member and on the tab for other san devices. The issues impacting the display of multipath devices in the user interface were also contributing to an improper count of selected and total devices in the user interface. Added a data structure to track tallied storage devices and account for the movement of multipath devices from single other san devices over to the multipath tab. Related: rhbz#1014425 Resolves: rhbz#914637 --- iw/filter_gui.py | 154 +++++++++++++++++++++++++++++++++++++++---------------- 1 file changed, 111 insertions(+), 43 deletions(-) diff --git a/iw/filter_gui.py b/iw/filter_gui.py index 6c6f0df..ae865ef 100644 --- a/iw/filter_gui.py +++ b/iw/filter_gui.py @@ -99,6 +99,7 @@ def insert(self, index, value): selectedDevices = 0 totalSize = 0 selectedSize = 0 +talliedDevices =[] # These are global so they can be accessed from all Callback objects. The # basic callback defines its membership as anything that doesn't pass the @@ -109,7 +110,6 @@ def isCCISS(info): def isRAID(info): if flags.dmraid: return udev_device_is_biosraid(info) - return False def isMultipath(info): @@ -130,6 +130,24 @@ def __init__(self, xml): def addToUI(self, tuple): pass + def incrementDeviceCount(self, tuple): + global totalDevices, totalSize + global selectedDevices, selectedSize + global talliedDevices + + # Check to be sure this device has not already been counted + if tuple[0]["name"] in talliedDevices or tuple[0]["DEVNAME"] in talliedDevices: + return + + talliedDevices.append(tuple[0]["name"]) + + totalDevices += 1 + totalSize += tuple[0]["XXX_SIZE"] + + if tuple[2]: + selectedDevices += 1 + selectedSize += tuple[0]["XXX_SIZE"] + def deviceToggled(self, set, device): global selectedDevices, totalDevices global selectedSize, totalSize @@ -209,6 +227,31 @@ def addToUI(self, tuple): if not tuple[INTERCONNECT_COL] in self._interconnects: self._interconnects.append(tuple[INTERCONNECT_COL]) + def incrementDeviceCount(self, tuple): + global totalDevices, totalSize + global selectedDevices, selectedSize + global talliedDevices + + # Check to be sure this device has not already been counted + if tuple[0]["name"] in talliedDevices: + return + + # If the mpath member is already in the list, add the mpath device name + if tuple[0]["DEVNAME"] in talliedDevices: + talliedDevices.append(tuple[0]["name"]) + return + + # If neither the mpath device nor the mpath member exist, add both and increment + talliedDevices.append(tuple[0]["name"]) + talliedDevices.append(tuple[0]["DEVNAME"]) + + totalDevices += 1 + totalSize += tuple[0]["XXX_SIZE"] + + if tuple[2]: + selectedDevices += 1 + selectedSize += tuple[0]["XXX_SIZE"] + def isMember(self, info): return info and isMultipath(info) @@ -285,7 +328,7 @@ def __init__(self, *args, **kwargs): self.IDEntry.connect("changed", lambda entry: self.model.get_model().refilter()) def isMember(self, info): - return info and isOther(info) + return info and isOther(info) and not isMultipath(info) class SearchCallbacks(FilteredCallbacks): def __init__(self, *args, **kwargs): @@ -310,6 +353,10 @@ def __init__(self, *args, **kwargs): self.LUNEntry.connect("changed", lambda entry: self.model.get_model().refilter()) self.IDEntry.connect("changed", lambda entry: self.model.get_model().refilter()) + # Search window shows all devices, do not let it double count them + def incrementDeviceCount(self, tuple): + return + def isMember(self, info): return True @@ -457,51 +504,78 @@ def _add_advanced_clicked(self, button): writeMultipathConf(friendly_names=True) if os.path.exists("/etc/multipath.conf"): - (new_singlepaths, new_mpaths, new_partitions) = identifyMultipaths(new_disks) + (new_singlepaths, mpath_conf, new_partitions) = identifyMultipaths(new_disks) else: new_singlepaths = new_disks - new_mpaths = [] + mpath_conf = [] (new_raids, new_nonraids) = self.split_list(lambda d: isRAID(d) and not isCCISS(d), new_singlepaths) - # The end result of the loop below is that mpaths is a list of lists of - # components, just like new_mpaths. That's what populate expects. + # mpaths is a list of multipath devices where each multipath device is represented + # by a list of udev dictionaries which each describe a member of the mpath device + # EG: [[{sda_udev_dict}, {sdb_udev_dict}], [{sdd_udev_dict}, {sde_udev_dict}]] + # This is the data structure expected by the populate method mpaths = [] - for mp in new_mpaths: + # mpath_conf: a data structure that is a list of lists describing the mpath + # devices from the multipath config file; structure is equivalent + # to that of mpaths shown above + # mpath_device: represents one element from mpath_conf whose own list + # elements are udev dictionaries describing the members of + # multipath devices in the multipath config file + for mpath_device in mpath_conf: old_mpaths = [] is_new_multipath = False - for d in mp: + # mpath_member: a udev dictionary describing a block device that is a + # member of a configured mpath device + for mpath_member in mpath_device: # If any of the multipath components are in the nonraids cache, # remove them from the UI store. - if d in self._cachedDevices: - self.depopulate(d) - - # Check whether this device is part of an existing multipath, - # but as part of a different set of members - if d in self._cachedMPathMembers and hasattr(self._cachedMPaths, d["name"]): - if self._cachedMPaths[d["name"]] != mp: - old_mpaths.append(mp) + if mpath_member in self._cachedDevices: + self.depopulate(mpath_member) + + # self._cachedMPathMembers (subclass of the NameCache class) + # list of udev dictionaries for each block device that is a member + # of a multipath device + # [ {udev_dict}, {udev_dict}, {udev_dict}] + # The NameCache class is a custom created "magic class" whose internals + # change the syntax for lookups and iteration. + + # self._cachedMPaths + # dictionary that contains an entry for each block device shown + # in the UI before the current multipath config file was read + # { sdX :[{udev_dict}, {udev_dict}], sdY: [{udev_dict}, {udev_dict}]} + + # Check for the member device in the UI caches + if mpath_member in self._cachedMPathMembers and \ + mpath_member["name"] in self._cachedMPaths: + # Check if the former multipath device has changed and has not + # been added to the old_mpaths list + if self._cachedMPaths[mpath_member["name"]] != mpath_device and \ + self._cachedMPaths[mpath_member["name"]] not in old_mpaths: + old_mpaths.append(self._cachedMPaths[mpath_member["name"]]) is_new_multipath = True # Check whether this is a new device - if d not in self._cachedMPathMembers: + if mpath_member not in self._cachedMPathMembers: is_new_multipath = True - for omp in old_mpaths: - # Clean up the caches - for d in omp: - del self._cachedMPathMembers[self._cachedMPathMembers.index(d)] - del self._cachedMPaths[d["name"]] - self.depopulate(d) + # Remove multipath devices from the UI caches if they changed in the + # multipath config file that was read into mpath_conf + for cached_mpath_device in old_mpaths: + # Check for each member device in the caches + for mpath_member in cached_mpath_device: + del self._cachedMPathMembers[self._cachedMPathMembers.index(mpath_member["name"])] + del self._cachedMPaths[mpath_member["name"]] + self.depopulate(mpath_member) if is_new_multipath: - mpaths.append(mp) + mpaths.append(mpath_device) # Add the multipath to the cache - for d in mp: - self._cachedMPaths[d["name"]] = mp + for mpath_member in mpath_device: + self._cachedMPaths[mpath_member["name"]] = mpath_device nonraids = filter(lambda d: d not in self._cachedDevices, new_nonraids) raids = filter(lambda d: d not in self._cachedRaidDevices, new_raids) @@ -514,8 +588,7 @@ def _add_advanced_clicked(self, button): self._cachedDevices.extend(nonraids) self._cachedRaidDevices.extend(raids) - # And then we need to do the same list flattening trick here as in - # getScreen. + # And then we need to do the same list flattening trick here as in getScreen. lst = list(itertools.chain(*mpaths)) self._cachedMPathMembers.extend(lst) @@ -703,32 +776,24 @@ def depopulate(self, component): for row in self.store: if row[4] == component['DEVNAME']: self.store.remove(row.iter) - return + # Also check for mpath devices to depopulate if they have been updated + elif 'ID_MPATH_NAME' in component and row[4] == component['ID_MPATH_NAME']: + self.store.remove(row.iter) + return def populate(self, nonraids, mpaths, raids, activeByDefault=False): def _addTuple(tuple): global totalDevices, totalSize global selectedDevices, selectedSize - added = False self.store.append(None, tuple) + # for each page in the notebook, check to see if each existing device should + # be added to the user interface and added to the device tally for pg in self.pages: if pg.cb.isMember(tuple[0]): - added = True pg.cb.addToUI(tuple) - - # Only update the size label if this device was added to any pages. - # This prevents situations where we're only displaying the basic - # filter that has one disk, but there are several advanced disks - # in the store that cannot be seen. - if added: - totalDevices += 1 - totalSize += tuple[0]["XXX_SIZE"] - - if tuple[2]: - selectedDevices += 1 - selectedSize += tuple[0]["XXX_SIZE"] + pg.cb.incrementDeviceCount(tuple) def _isProtected(info): protectedNames = map(udev_resolve_devspec, self.anaconda.protected) @@ -756,6 +821,8 @@ def _active(info): else: return False + # Entry point to the code that runs when the FilterWindow GUI is opened + # within the anaconda installer for d in nonraids: name = udev_device_get_name(d) @@ -844,6 +911,7 @@ def _active(info): # We use a copy here, so as to not modify the original udev info # dict as that would break NameCache matching data = mpath[0].copy() + # This will swap the device name (eg: sda) for the mpath name (eg: mpatha) data["name"] = udev_device_get_multipath_name(mpath[0]) tuple = (data, True, _active(data), _isProtected(data), udev_device_get_multipath_name(mpath[0]), model, -- To view this commit on github, visit https://github.com/rhinstaller/anaconda/commit/4843f8165e6089fc0c8ecc495f1e…

1 0

New: [rhinstaller/anaconda/pulls/475 rhel6-branch] Fix multipath UI representation (#914637)
by dashea 18 Dec '15

18 Dec '15

Changed the depopulate method of the FilterWindow class to correctly identify and remove outdated representations of a multipath device in the user interface. Added a check to ensure the first member device of a multipath device is not accidentally displayed and counted as both an mpath device member and on the tab for other san devices. The issues impacting the display of multipath devices in the user interface were also contributing to an improper count of selected and total devices in the user interface. Added a data structure to track tallied storage devices and account for the movement of multipath devices from single other san devices over to the multipath tab. Related: rhbz#1014425 Resolves: rhbz#914637 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/475

1 0

[PATCH rhel6-branch] fcoe: take into account vlan device names trimmed to IFNAMSIZ (#1021067)
by Radek Vykydal 18 Dec '15

18 Dec '15

Resolves: rhbz#1021067 --- storage/udev.py | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/storage/udev.py b/storage/udev.py index 78fb996..cd433c3 100644 --- a/storage/udev.py +++ b/storage/udev.py @@ -620,6 +620,10 @@ def udev_device_get_iscsi_initiator(info): # and find whether the host has 'fc_host' and if it the device has a bound # Ethernet interface. +# bug #1021067 - due to IFNAMESIZ limit, device name can be trimmed +def _fcoe_suffix_in_sysfs_path(sysfs_path): + return any(fcoe_suffix in sysfs_path for fcoe_suffix in ["-fcoe", "-fco", "-fc", "-f"]) + def _detect_broadcom_fcoe(info): re_pci_host=re.compile('/(.*)/(host\d+)') match = re_pci_host.match(info["sysfs_path"]) @@ -641,7 +645,7 @@ def udev_device_is_fcoe(info): path_components[2] == "fc": return True - if path.startswith("fc-") and "fcoe" in info["sysfs_path"]: + if path.startswith("fc-") and _fcoe_suffix_in_sysfs_path(info["sysfs_path"]): return True if _detect_broadcom_fcoe(info) != (None, None): @@ -657,7 +661,7 @@ def udev_device_get_fcoe_nic(info): path_components[2] == "fc": return path_components[1] - if path.startswith("fc-") and "fcoe" in info["sysfs_path"]: + if path.startswith("fc-") and _fcoe_suffix_in_sysfs_path(info["sysfs_path"]): return info["sysfs_path"].split("/")[4].split(".")[0] (sysfs_pci, host) = _detect_broadcom_fcoe(info) @@ -679,7 +683,7 @@ def udev_device_get_fcoe_identifier(info): path_components[2] == "fc": return path_components[3] - if path.startswith("fc-") and "fcoe" in info["sysfs_path"]: + if path.startswith("fc-") and _fcoe_suffix_in_sysfs_path(info["sysfs_path"]): return path_components[1] if udev_device_is_fcoe(info) and len(path_components) >= 4 and \ -- 2.4.3

2 1

Re: [rhinstaller/lorax/pulls/77 master] Write a list of debuginfo packages to /root/debug-pkgs.log (#1068675)
by dashea 18 Dec '15

18 Dec '15

I understand the gdb remote case, but one can easily query for debuginfo packages post-facto given the ENVR from the rpmdb. Having the full rpmdb would have helped me in the past for other things as well, mostly knowing the version of `python-kickstart` etc. So it's not "no useful reason". One other minor negative here is it's another copy+paste of some historical lore around the naming of debuginfo repositories. As for the size...I'm somewhat sympathetic, but the biggest win in this area would be a headless boot.iso for people who want to PXE + kickstart. For live or other media, it's already huge. -- To view this pull request on github, visit https://github.com/rhinstaller/lorax/pull/77

1 0

Re: [rhinstaller/anaconda/pulls/474 rhel7-branch] rpmostreepayload: Only add remote if it doesn't exist
by dashea 18 Dec '15

18 Dec '15

The code in master uses the API with `OSTree.RepoRemoteChange.ADD_IF_NOT_EXISTS`, so should be fine. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/474

1 0

← Newer
1
...
4
5
6
7
8
9
10
...
50
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

anaconda-patches December 2015