[ovirt-users] stange behavious on virtual machine wilt multiple snapshots

[Roel de Rooy]

I managed to stop the failed messages from being logged within the event log.
I used the clear async task option within vdsClient  to remove the task from the host.

After restarting the engine daemon, the error message within the events and the finalizing subtasks stopped.
The task itself is still visible as running, so my best guess is that  I have to manually remove it from the database, correct?
Already updated the specific tables within the database to unlock the VM and the disk inside VM.

Unfortunately I am still puzzled on how to proceed with the error message the VM is producing:

"down with error. Exit message: internal error: early end of file from monitor: possible problem:
2015-10-20T00:06:08.134352Z qemu-system-x86_64: -drive file=/rhev/data-center/124ae76f-8acb-412e-91cc-dff9f6ec665d/7be6092d-fce1-45b5-a383-3ead3cdfe457/images/bb5d12ce-1845-42c8-8a6c-abbb73999f5d/941ba152-9bc7-4aaa-8dfa-f1b9f7c3281b,if=none,id=drive-virtio-disk0,format=qcow2,serial=bb5d12ce-1845-42c8-8a6c-abbb73999f5d,cache=none,werror=stop,rerror=stop,aio=threads: could not open disk image /rhev/data-center/124ae76f-8acb-412e-91cc-dff9f6ec665d/7be6092d-fce1-45b5-a383-3ead3cdfe457/images/bb5d12ce-1845-42c8-8a6c-abbb73999f5d/941ba152-9bc7-4aaa-8dfa-f1b9f7c3281b: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file."


When looking within the vdsm log; before the attempt is made to start the VM, the following messages are visible for all images within the specific VM folder:
"Thread-670750::DEBUG::2015-10-21 10:15:12,775::fileSD::499::Storage.StorageDomain::(activateVolumes) Fixing permissions on /rhev/data-center/mnt/ ."

Could this be permissions related?

Kind regards,
Roel


Van: users-bounces at ovirt.org [mailto:users-bounces at ovirt.org] Namens Roel de Rooy
Verzonden: 21 October 2015 09:45
Aan: users at ovirt.org
Onderwerp: [ovirt-users] stange behavious on virtual machine wilt multiple snapshots

Hi all,

One of our VM's is behaving strangely.
We are using oVirt Engine 3.5.4.
The host where the VM is located is F20 with virt-preview repository installed.

When trying to start the VM, the following error message is visible:

"down with error. Exit message: internal error: early end of file from monitor: possible problem:
2015-10-20T00:06:08.134352Z qemu-system-x86_64: -drive file=/rhev/data-center/124ae76f-8acb-412e-91cc-dff9f6ec665d/7be6092d-fce1-45b5-a383-3ead3cdfe457/images/bb5d12ce-1845-42c8-8a6c-abbb73999f5d/941ba152-9bc7-4aaa-8dfa-f1b9f7c3281b,if=none,id=drive-virtio-disk0,format=qcow2,serial=bb5d12ce-1845-42c8-8a6c-abbb73999f5d,cache=none,werror=stop,rerror=stop,aio=threads: could not open disk image /rhev/data-center/124ae76f-8acb-412e-91cc-dff9f6ec665d/7be6092d-fce1-45b5-a383-3ead3cdfe457/images/bb5d12ce-1845-42c8-8a6c-abbb73999f5d/941ba152-9bc7-4aaa-8dfa-f1b9f7c3281b: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file: Could not open backing file."

After investigating this issue, we found out that the VM had 25 disk snapshots attached.
We tried to remove 2 of them, which succeeded, but due to the time it would take to remove them all, we figured it would also be possible to just export the vm to the exportdisk and reimport it again.

Unfortunately now the attached screenshot is happening.

Each of the 'finalizing' subtasks corresponds with a 'Failed to export Vm xxxx to xxxx' message within the event log.

Is there a possibility  to stop this task?

The vdsClient on the host gives the following info about the task:

5c2763ce-21b7-4669-b66d-5f5ea075be4a :
        verb = moveImage
        code = 0
        state = finished
        tag = spm
        result =
        message = 1 jobs completed successfully
        id = 5c2763ce-21b7-4669-b66d-5f5ea075be4a