- Devel - libvirt List Archives

[libvirt] image locking?
by Itamar Heim 22 Oct '08

22 Oct '08

Hi, I am interested to find out how libvirt envisions image locking. i.e., how do we make sure multiple nodes are not trying to access the same storage volume, probably causing image corruption. I know this can be solved by means of a cluster, but it seems excessive (and not possible in all scenarios). Central administration (ovirt-like) is also problematic, since unless fencing is assumed, it cannot verify the image is no longer being used. If the image/storage volume could be locked (or leased lock), either by the storage/LVM/NFS preventing from a node access to a specific image, or by having libvirt/qemu mutually check the lock before accessing it (very simply, in leased lock we can check every X minutes, and kill the process to make sure it honors the lock). Any thoughts on the subject? Thanks, Itamar

4 11

[libvirt] errors building rpm
by Ben Guthro 22 Oct '08

22 Oct '08

I keep getting the following error when trying to build via a 'sudo make rpm' on the tip of the tree + /usr/lib/rpm/find-lang.sh /var/tmp/libvirt-0.4.6-1.fc9-root libvirt No translations found for libvirt in /var/tmp/libvirt-0.4.6-1.fc9-root error: Bad exit status from /var/tmp/rpm-tmp.82848 (%install) RPM build errors: Bad exit status from /var/tmp/rpm-tmp.82848 (%install) make: *** [rpm] Error 1 This looks like it came in with one of Jim Meyering's latest checkins. I'm wondering if I'm doing something wrong - or if I'm missing a build dependency that I did not need previously? Ben Guthro

5 6

[libvirt] [PATCH 00/12] Domain Events - Round 3
by Ben Guthro 21 Oct '08

21 Oct '08

The following patch series implements the Events API discussed here previously in this thread: http://www.redhat.com/archives/libvir-list/2008-September/msg00321.html http://www.redhat.com/archives/libvir-list/2008-October/msg00245.html and https://www.redhat.com/archives/libvir-list/2008-October/msg00432.html Once again, I have broken these into the following patches: events-01-public-api Changes to libvirt.h, and libvirt.c events-02-internal-api Changes to internal.h and event code events-03-qemud Changes to the daemon events-04-qemud-rpc-protocol Changes to the qemud rpc protocol events-05-driver.patch Changes to the driver API events-06-driver-qemu Changes to the qemu driver events-07-driver-remote Changes to the remote driver events-08-driver-lxc Minor changes to LXC driver structure to prevent compile errors events-09-driver-openvz Minor changes to openvz driver structure to prevent compile errors events-10-driver-test Minor changes to test driver structure to prevent compile errors events-11-example-testapp Test app, and infrastructure changes for an example dir events-12-python-ignore.patch Add functions to be ignored, for now Differences from last submission - Formatting issues brought up by Daniel V Some protocol simplifications General fixes suggested from last time around. Removal of POLLxxx stuff (this caused a bit of churn) I think I got them all, but please let me know if I missed something. Again, domain events are only emitted from qemu/kvm guests. Ben Guthro

1 0

[libvirt] [PATCH 0 of 2] Cgroup and LXC fixes
by Dan Smith 21 Oct '08

21 Oct '08

This set moves the cgroup creation before that of the container process, thus ensuring that it is put in the cgroup as well. As a result, I noticed that we need to allow device access to /dev/pts/*, and thus added a cgroup mechanism to allow a whole major device type. The LXC driver is made to allow major type 136 as a result. Note that this doesn't seem to do much to really restrict the container. While it does prevent them from opening devices other than what are allowed, the container can still mount (or access) the cgroup filesystem and move itself out of its own group and into the unrestricted root. Further, it can just add whitelist entries for the devices it wants to gain access. I tested code to restrict the devices in the per-driver cgroup, but that appears to have no effect, because from within the container, I can still add "b 8:* rwm" to my group's devices.allow and subsequently access SCSI disks. Even still, this patch set is crucial for proper cgroup membership of the container children.

3 9

[libvirt] [PATCH 3/5]: Set format type to LVM2 on NULL format
by Chris Lalancette 21 Oct '08

21 Oct '08

Again because of my generic VIR_ENUM_IMPL patch last week, I unwittingly caused a regression in the storage_backend_logical driver. Previously, if you submitted logical pool XML that had no <source><format type='lvm2'/> string, it would just default to VIR_STORAGE_POOL_LOGICAL_LVM2. This would succeed just fine and go on with life. This is no longer the case, and now XML without the format tag fails to define. To maintain backwards compatibility with already existing XML that expects this, add a compatibility wrapper to return VIR_STORAGE_POOL_LOGICAL_LVM2 on an empty format tag. Signed-off-by: Chris Lalancette <clalance(a)redhat.com>

2 2

[libvirt] [PATCH 4/5]: Make dumpXML more idempotent with defineXML
by Chris Lalancette 21 Oct '08

21 Oct '08

Currently, you can define a logical storage pool with something like the following: <pool type='logical'> <source> <name>MyVG</name> <device name='/dev/sdb'/> ... However, dumping out the XML for this same storage pool (with, say, virsh pool-dumpxml), gives: <pool type='logical'> <source> <name>MyVG</name> <device name='/dev/sdb'> </device> To make this more idempotent, do the <device name='/dev/sdb'/> form by default, and only do the <device>...</device> form if .nfreeExtent is defined for the storage pool. Signed-off-by: Chris Lalancette <clalance(a)redhat.com>

2 2

[libvirt] [PATCH 2/5]: Properly set errors on unknown format types
by Chris Lalancette 21 Oct '08

21 Oct '08

Because of my patch last week that converted the various virStorage*FromString and virStorage*ToString implementations to the generic VIR_ENUM_IMPL, there were a couple of places that didn't properly set errors when they failed. This patch fixes these places up. It also adds an additional check in virEnumFromString(), so that if a NULL type is passed in, it fails instead of SEGV'ing. Signed-off-by: Chris Lalancette <clalance(a)redhat.com>

2 2

[libvirt] [PATCH 1/5]: Shore up virGetLastError handling
by Chris Lalancette 21 Oct '08

21 Oct '08

While doing testing on this patch series, I mistakenly added a bogus piece of storage XML to libvirtd, which was saved in /etc/libvirt/storage. On subsequent stop/start of libvirtd, because of another bug (fixed in a later patch), an error wasn't being set properly in an error path, so libvirtd was SEGV'ing in storage_conf.c:virStoragePoolObjLoad when trying to dereference the NULL err returned from virGetLastError(). Make this more robust against errors by always doing "err ? err->message : NULL" in the printf. I looked around the tree and found a couple of other places that weren't guarded, so this patch fixes them as well. Signed-off-by: Chris Lalancette <clalance(a)redhat.com>

2 2

[libvirt] [PATCH] minor usb masstorage hotadd cleanup
by Guido Günther 21 Oct '08

21 Oct '08

Hi, this just brings qemudDomainAttachUsbMassstorageDevice in line with the rest of the code: * don't allow to add the same target device name more than once * escape paths for qemu's monitor command Cheers, -- Guido

2 2

[libvirt] [PATCH 00/12] Domain Events - Round 2
by Ben Guthro 21 Oct '08

21 Oct '08

The following patch series implements the Events API discussed here previously in this thread: http://www.redhat.com/archives/libvir-list/2008-September/msg00321.html and http://www.redhat.com/archives/libvir-list/2008-October/msg00245.html By Daniel B's advice - I have broken this patch into the following: events-01-public-api Changes to libvirt.h, and libvirt.c events-02-internal-api Changes to internal.h and event code events-03-qemud Changes to the daemon events-04-qemud-rpc-protocol Changes to the qemud rpc protocol events-05-driver.patch Changes to the driver API events-06-driver-qemu Changes to the qemu driver events-07-driver-remote Changes to the remote driver events-08-driver-lxc Minor changes to LXC driver structure to prevent compile errors events-09-driver-openvz Minor changes to openvz driver structure to prevent compile errors events-10-driver-test Minor changes to test driver structure to prevent compile errors events-11-example-testapp Test app, and infrastructure changes for an example dir events-12-python-ignore.patch Add functions to be ignored, for now For now, domain events are only emitted from qemu/kvm guests.

5 31