[pve-devel] VM locked after failed Snapshot

[Stefan Priebe - Profihost AG]

Am 09.09.2014 um 07:52 schrieb Dietmar Maurer:
>> Am 08.09.2014 17:11, schrieb Dietmar Maurer:
>>> BTW, is there a way to reproduce the bug?
>>
>> Still not sure if it is a bug or a feature ;-)
>>
>> Just interrupt the rbd process while doing a snapshot or the pve worker.
> 
>>>      if ($err) {
>>>          warn "snapshot create failed: starting cleanup\n";
>>>          eval { snapshot_delete($vmid, $snapname, 0, $drivehash); };
> 
> I think snapshot delete should work in that case. Can we modify it to make it work
> for that situation?

We generally need to check whether deletion mentions watchers - then
pause 30s and retry.

But i think this will be very ugly code...

Stefan

>>>          warn $@ if $@;
>>>          die $err;
>>>      }
> 
>