[ovirt-users] Cloning VM on NFS Leads to Locked Disks

Fred Rolland frolland at redhat.com
Sun Jul 2 14:32:19 UTC 2017


Seems you hit [1].
Can you tell me what is the exact version of the engine ?

[1] https://bugzilla.redhat.com/show_bug.cgi?id=1395793

On Thu, May 11, 2017 at 10:29 PM, Charles Tassell <charles at islandadmin.ca>
wrote:

> Sure, it's pretty big so I've put it online for download at
> http://krissy.islandadmin.ca/public/engine.log.txt
>
>
> On 2017-05-11 04:08 PM, Fred Rolland wrote:
>
> The locking is on the engine side and restarting the vdsm will not affect
> it .
> Can you send the whole engine log ?
> Which exact version are you using ?
>
>
> On Thu, May 11, 2017 at 9:30 PM, Charles Tassell <charles at islandadmin.ca>
> wrote:
>
>> Just as an update, I created a new VM and had the same issue: the disk
>> remains locked.  So I then added a new data store (this one iSCSI not NFS)
>> and create a new VM on that.  Again, the disk remains locked.  So the
>> problem seems to be that any action that sets to modify a disk image on my
>> cluster locks the disk and keeps it locked permanently.
>>
>> I tried restarting the vdsm daemon, but that didn't make a difference.
>> I'm seeing this in my sanlock.log file though, which worries me:
>>
>>
>> 017-05-07 07:51:41-0300 1738538 [13575]: s2 renewal error -202
>> delta_length 10 last_success 1738508
>> 2017-05-07 07:51:41-0300 1738538 [11513]: s1 renewal error -202
>> delta_length 10 last_success 1738508
>>
>> Here's the last 20 lines:
>> 2017-05-07 07:51:41-0300 1738538 [13580]: s3 renewal error -202
>> delta_length 10 last_success 1738508
>> 2017-05-07 07:51:41-0300 1738538 [13575]: 20423d5e aio timeout RD
>> 0x7fe1440008c0:0x7fe1440008d0:0x7fe160255000 ioto 10 to_count 67
>> 2017-05-07 07:51:41-0300 1738538 [13575]: s2 delta_renew read timeout 10
>> sec offset 0 /rhev/data-center/mnt/192.168.130.217:_media_ovirt/20423d5e-
>> 188c-4e10-9893-588ceb81b354/dom_md/ids
>> 2017-05-07 07:51:41-0300 1738538 [13575]: s2 renewal error -202
>> delta_length 10 last_success 1738508
>> 2017-05-07 07:51:41-0300 1738538 [11513]: hosted-e aio timeout RD
>> 0x7fe1480008c0:0x7fe1480008d0:0x7fe14e6fc000 ioto 10 to_count 65
>> 2017-05-07 07:51:41-0300 1738538 [11513]: s1 delta_renew read timeout 10
>> sec offset 0 /var/run/vdsm/storage/5dccd07d-a923-4d4b-9cb1-3b51ebfdca4d/
>> 5a9c284f-0faa-4a25-94ce-c9efdae07484/ab2443f1-95ed-475d-886c-c1653257cf04
>> 2017-05-07 07:51:41-0300 1738538 [11513]: s1 renewal error -202
>> delta_length 10 last_success 1738508
>> 2017-05-07 07:51:47-0300 1738544 [13575]: 20423d5e aio collect RD
>> 0x7fe1440008c0:0x7fe1440008d0:0x7fe160255000 result 1048576:0 match reap
>> 2017-05-07 07:51:47-0300 1738544 [13580]: 5dccd07d aio collect RD
>> 0x7fe13c0008c0:0x7fe13c0008d0:0x7fe14e5fa000 result 1048576:0 match reap
>> 2017-05-07 07:51:47-0300 1738544 [11513]: hosted-e aio collect RD
>> 0x7fe1480008c0:0x7fe1480008d0:0x7fe14e6fc000 result 1048576:0 match reap
>> 2017-05-07 07:53:57-0300 1738674 [13590]: s2:r15 resource
>> 20423d5e-188c-4e10-9893-588ceb81b354:SDM:/rhev/data-center/
>> mnt/192.168.130.217:_media_ovirt/20423d5e-188c-4e10-9893-
>> 588ceb81b354/dom_md/leases:1048576 for 7,21,78395
>> 2017-05-07 07:59:49-0300 1739027 [13575]: s2 delta_renew long write time
>> 10 sec
>> 2017-05-09 08:38:34-0300 1914151 [13590]: s2:r16 resource
>> 20423d5e-188c-4e10-9893-588ceb81b354:SDM:/rhev/data-center/
>> mnt/192.168.130.217:_media_ovirt/20423d5e-188c-4e10-9893-
>> 588ceb81b354/dom_md/leases:1048576 for 7,21,78395
>> 2017-05-11 15:07:45-0300 2110302 [13590]: s2:r17 resource
>> 20423d5e-188c-4e10-9893-588ceb81b354:SDM:/rhev/data-center/
>> mnt/192.168.130.217:_media_ovirt/20423d5e-188c-4e10-9893-
>> 588ceb81b354/dom_md/leases:1048576 for 7,21,112346
>> 2017-05-11 15:17:24-0300 2110881 [13590]: s4 lockspace
>> b010093e-1924-46e1-bd57-2cf2b2445087:1:/dev/b010093e-1924-
>> 46e1-bd57-2cf2b2445087/ids:0
>> 2017-05-11 15:17:45-0300 2110902 [1395]: s4 host 1 1 2110881
>> 44ae07eb-3371-4750-8728-ab3b049dbae2.ovirt730-0
>> 2017-05-11 15:17:45-0300 2110902 [1400]: s4:r18 resource
>> b010093e-1924-46e1-bd57-2cf2b2445087:SDM:/dev/b010093e-1924-
>> 46e1-bd57-2cf2b2445087/leases:1048576 for 7,21,112346
>> 2017-05-11 15:17:52-0300 2110909 [1399]: s5 lockspace
>> b010093e-1924-46e1-bd57-2cf2b2445087:1:/dev/b010093e-1924-
>> 46e1-bd57-2cf2b2445087/ids:0
>> 2017-05-11 15:18:13-0300 2110930 [1395]: s5 host 1 2 2110909
>> 44ae07eb-3371-4750-8728-ab3b049dbae2.ovirt730-0
>> 2017-05-11 15:18:13-0300 2110930 [1395]: s5 host 2 1 2110065
>> 4d31313f-b2dd-4368-bf31-d39835e10afb.ovirt730-0
>>
>>
>> On 2017-05-11 10:09 AM, Charles Tassell wrote:
>>
>> Hi Freddy,
>>
>>   Sure, thanks for looking into this.  Here you go:
>>
>> 2017-05-10 11:35:30,249-03 INFO  [org.ovirt.engine.core.bll.aaa.SessionDataContainer]
>> (DefaultQuartzScheduler8) [1c84abac] Not removing session
>> 'vZyqrcCljPC7hQtcILsk4uDug3QsiinZBOyoGDiQKkYYT2znGyWe4fasrPb
>> jYxdjbfyR3DBnp+UZ9/k20dGsMA==', session has running commands for user
>> 'admin at internal-authz'.[snip]
>> On 2017-05-11 04:30 AM, Fred Rolland wrote:
>>
>> Hi,
>>
>> Can you provide the engine log ?
>>
>> Thanks,
>>
>> Freddy
>>
>> On Wed, May 10, 2017 at 5:57 PM, Charles Tassell <charles at islandadmin.ca>
>> wrote:
>>
>>> Hi Everyone,
>>>
>>>   I'm having some issues with my oVirt 4.1 (fully updated to latest
>>> release as of yesterday) cluster.  When I clone a VM the disks of both the
>>> original and the clone stay in the locked state, and the only way I can
>>> resolve it is to go into the database on the engine and run "update images
>>> set imagestatus=1 where imagestatus=2;"
>>>
>>>   I'm using NFS4 as a datastore and the disks seem to copy fine (file
>>> sizes match and everything), but the locking worries me.  To clone the VM I
>>> just shut the source VM down and then right click on it and select "Clone"
>>>
>>>   I've attached the full VDSM log from my last attempt, but here is the
>>> excerpt of the lines just referencing the two disks
>>> (d73206ed-89ba-48a9-82ff-c107c1af60f0 is the original VMs disk and
>>> 670a7b20-fecd-45c6-af5c-3c7b98258224 is the clone.)
>>>
>>> [snip]
>>> _______________________________________________
>>> Users mailing list
>>> Users at ovirt.org
>>> http://lists.ovirt.org/mailman/listinfo/users
>>>
>>>
>>
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ovirt.org/pipermail/users/attachments/20170702/e0c7f5aa/attachment-0001.html>


More information about the Users mailing list