[BLOCKED] vdsm: [Scalability] when performing several storage related tasks, creation of live snapshots takes over 5 minutes in vdsm and dies in engine because of timeout
Created attachment 649220[details]
logs
Description of problem:
I ran 20 live snapshot + 2 deleteImage with wipe after delete and we are getting a broken pipe error.
Version-Release number of selected component (if applicable):
vdsm-4.9.6-44.0.el6_3.x86_64
libvirt-0.9.10-21.el6_3.6.x86_64
How reproducible:
100%
Steps to Reproduce:
1. create 20 thin provision vms and run them (make sure they are writing)
2. remove 20GB and a 10GB preallocated disks with wipe after delete
3. create live snapshots for each of the vms
Actual results:
some of the snapshots fail to be created because vdsm takes more than 5 minutes to create the snapshot and engine gets timeout for the tasks.
Expected results:
user should be able t run multiple actions without timeout
PLEASE NOTE: although we get a failure in the create snapshot UI is showing the snapshot as created. this might cause problem for user in chain.
Additional info:logs
(In reply to comment #2)
> Trying to understand the threshold that kills the snapshot.
> If we do not perform the deletes and wipe-after-delete, will the snapshot
> still fail?
it might, but not with timeout error.
the problem is that wipe after delete takes a long time.
since we have a queue for tasks, and if we have a long running task, other tasks that are started will get timeout.
This is an automated message.
This Bugzilla report has been opened on a version which is not maintained anymore.
Please check if this bug is still relevant in oVirt 3.5.4.
If it's not relevant anymore, please close it (you may use EOL or CURRENT RELEASE resolution)
If it's an RFE please update the version to 4.0 if still relevant.
Comment 10Sandro Bonazzola
2015-10-02 11:02:13 UTC
This is an automated message.
This Bugzilla report has been opened on a version which is not maintained
anymore.
Please check if this bug is still relevant in oVirt 3.5.4 and reopen if still
an issue.
Created attachment 649220 [details] logs Description of problem: I ran 20 live snapshot + 2 deleteImage with wipe after delete and we are getting a broken pipe error. Version-Release number of selected component (if applicable): vdsm-4.9.6-44.0.el6_3.x86_64 libvirt-0.9.10-21.el6_3.6.x86_64 How reproducible: 100% Steps to Reproduce: 1. create 20 thin provision vms and run them (make sure they are writing) 2. remove 20GB and a 10GB preallocated disks with wipe after delete 3. create live snapshots for each of the vms Actual results: some of the snapshots fail to be created because vdsm takes more than 5 minutes to create the snapshot and engine gets timeout for the tasks. Expected results: user should be able t run multiple actions without timeout PLEASE NOTE: although we get a failure in the create snapshot UI is showing the snapshot as created. this might cause problem for user in chain. Additional info:logs