Red Hat Bugzilla – Bug 1257529
job monitoring don't work as expected
Last modified: 2016-04-19 21:39:41 EDT
Created attachment 1067616 [details]
Description of problem:
In many cases the job monitoring "reset" jobs that already finished and move them to status started - See this behavior via the UI.
In DB all seems to be fine (All jobs are finished)
* The most common job is when moving disk to other SD
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. create vm woth disk
2. Move the disk to second SD
In the logs - around 11:57:42
Another jobs is removing snapshot
Moti, I guess it is the same issue we're working on.
(In reply to Oved Ourfali from comment #2)
> Moti, I guess it is the same issue we're working on.
Indeed, this was solved by  and is another aspect of Bug 1248090
job 'Removing Snapshot .* of VM .*' still stuck
(In reply to ratamir from comment #1)
> Another jobs is removing snapshot
This scenario reproduced only for removing a snapshot of a running vm (with or without saving memory).
It seems that the Command Coordinator somehow nullify the job id associated with the action, which lead the command's context to fail in closing the job.
Verified on rhevm-188.8.131.52-0.1.el6.noarch (3.6.0-18)
Remove snapshot task marked as FINISHED after it completed