Bug 1298487 - Warn user when VMs with memory snapshots would end up in cluster with newer compatibility version
Summary: Warn user when VMs with memory snapshots would end up in cluster with newer c...
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: None
Product: ovirt-engine
Classification: oVirt
Component: BLL.Virt
Version: 3.6.0.3
Hardware: Unspecified
OS: Unspecified
high
high
Target Milestone: ovirt-3.6.3
: 3.6.3.1
Assignee: Marek Libra
QA Contact: Israel Pinto
URL:
Whiteboard:
Depends On: 1292398
Blocks: RHEV3.6Upgrade
TreeView+ depends on / blocked
 
Reported: 2016-01-14 09:17 UTC by Israel Pinto
Modified: 2016-03-11 07:23 UTC (History)
9 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2016-03-11 07:23:07 UTC
oVirt Team: Virt
Embargoed:
rule-engine: ovirt-3.6.z+
rule-engine: blocker+
mgoldboi: planning_ack+
michal.skrivanek: devel_ack+
mavital: testing_ack+


Attachments (Terms of Use)
engine_log_1 (197.10 KB, application/zip)
2016-01-14 09:17 UTC, Israel Pinto
no flags Details
engine_log_2 (1.62 MB, application/zip)
2016-01-14 09:18 UTC, Israel Pinto
no flags Details
vdsm_log (1.13 MB, application/zip)
2016-01-14 13:59 UTC, Israel Pinto
no flags Details
vm_qemu_log (3.12 KB, text/plain)
2016-01-14 14:33 UTC, Israel Pinto
no flags Details


Links
System ID Private Priority Status Summary Last Updated
oVirt gerrit 52859 0 master MERGED webadmin: Warn for suspended VMs when cluster change 2016-02-03 10:15:08 UTC
oVirt gerrit 53026 0 ovirt-engine-3.6 MERGED webadmin: Warn for suspended VMs when cluster change 2016-02-03 11:17:53 UTC
oVirt gerrit 53034 0 ovirt-engine-3.6.3 MERGED webadmin: Warn for suspended VMs when cluster change 2016-02-03 12:48:09 UTC

Description Israel Pinto 2016-01-14 09:17:30 UTC
Created attachment 1114733 [details]
engine_log_1

Description of problem:
After upgrade engine from 3.5.7 to 3.6.2 suspended VM failed to resume, it when down and up again.

Version-Release number of selected component (if applicable):
engine 3.5: 3.5.7-0.1.el6ev 
engine 3.6: 3.6.2-0.1.el6

Setup:
DC 3.5 with cluster 3.5
2 host 7.1 in cluster.
Steps to Reproduce:
1. Create VM and start it on Host_1
2. Suspend VM
3. Upgrade Host_2 and Host_1 to 7.2 
4. Upgrade Cluster level to 3.6
4. Start VM

Actual results:
VM is down, and restart again.

Expected results:
VM should run  and not got down

Additional info:
I see the error:
VM vm_71-2 is down with error. 
Exit message: Wake up from hibernation failed:internal error: cannot parse json {"return": , "id": "libvirt-99"}: parse error: unallowed token at this point in JSON text
                            {"return": , "id": "libvirt-99"}
                     (right here) ------^

Comment 1 Israel Pinto 2016-01-14 09:18:09 UTC
Created attachment 1114734 [details]
engine_log_2

Comment 2 Yaniv Kaul 2016-01-14 13:34:55 UTC
Israel, not sure but I think we'll need VDSM logs as well - please attach them.

Comment 3 Israel Pinto 2016-01-14 13:59:37 UTC
Created attachment 1114816 [details]
vdsm_log

Comment 4 Yaniv Kaul 2016-01-14 14:07:46 UTC
(In reply to Israel Pinto from comment #3)
> Created attachment 1114816 [details]
> vdsm_log

Thanks - much more helpful (though I could use third of it, no need for such a big log). If libvirt is in debug mode, then libvirt logs would be great too.

Comment 5 Israel Pinto 2016-01-14 14:33:53 UTC
Created attachment 1114820 [details]
vm_qemu_log

Comment 6 Israel Pinto 2016-01-14 14:34:41 UTC
(In reply to Yaniv Kaul from comment #4)
> (In reply to Israel Pinto from comment #3)
> > Created attachment 1114816 [details]
> > vdsm_log
> 
> Thanks - much more helpful (though I could use third of it, no need for such
> a big log). If libvirt is in debug mode, then libvirt logs would be great
> too.

it not in debug hope it can help to :)

Comment 7 Michal Skrivanek 2016-01-15 11:09:50 UTC
this is not planned to be supported. Suspended 3.5 VMs can't be resumed in 3.6 compatibility level. That's what the "compatibility level" means. Suspended VMs cannot be "upgraded" without losing their state, hence a forced Power Off is the only option

changing the scope of the bug to:
warn on cluster change
block resuming VM when compatibility level differs (careful - as master supports a custom one)

Comment 8 Michal Skrivanek 2016-01-15 11:49:31 UTC
actually, just the warning. The blocking is tracked under bug 1292398

Comment 9 Tomas Jelinek 2016-01-26 08:30:05 UTC
Changing title since the actual blocking of the resume is tracked in https://bugzilla.redhat.com/show_bug.cgi?id=1292398

Comment 10 Israel Pinto 2016-02-21 15:32:52 UTC
Verify:
engine 3.5: 3.5.7-0.1.el6ev
engine 3.6: 3.6.3.2-0.1.el6


Scenario:
1. Create VM and start it on Host_1
2. Suspend VM
3. Upgrade engine to 3.6 
4. Update VM cluster to 3.6 

Actual results as expected:
New window is opened with the content:
"Ti tel: Operation canceled
Error while executing action: 
Cannot update a VM in this status. Try stopping the VM first."

PASS


Note You need to log in before you can comment on or make changes to this bug.