This is a tracker bug for all bugs blocking the 3.4 release of oVirt.
May I ask to add BZ1110305 to the blocker list?
(In reply to Markus Stockhausen from comment #1)
weird, I'm not aware of any other setup with the same behavior. There might be something specific to your environment? is it a regression?
I guess the problem can only be seen in environments with heavy load. In our case the VM crashed during heavy disk IO. As the documentation attached to the BZ suggests other people have noticed the same BSODs . To mitigate the error the advice is to start Windows 7/2008/... VMs with relax option (to enable enlightment in the VM).
My understanding of this all is:
- We see a BSOD in a VM
- I open BZ1110305 for that to help to track the error down
- The advised solution is to add a qemu startup parameter
- So OVirt should set this parameter by default
What else should we do?
(In reply to Markus Stockhausen from comment #3)
BSODs are bad. Even though it is not a regression since 3.4 it is an important stability issue
(the other way around:)
BZ1113503 (although not being a regression) should be tested in 3.5 and if still present get a fix. At least in my opinion this improves stability (some kind of) too.
afaik BZ 1114914 should also be added as a blocker?
(In reply to Sven Kieske from comment #7)
> afaik BZ 1114914 should also be added as a blocker?
I agree, added.
as a blocker, as it is a regression which makes cloud-init unusable for custom scripts (e.g. set IPv6 configurations and everything cloud-init can't do out of the box)
Would you please add BZ1105513 as blocker since there isn't any workaround for it (afaik)?
(In reply to Stefano Stagnaro from comment #10)
> Would you please add BZ1105513 as blocker since there isn't any workaround
> for it (afaik)?
It's in modified state since 2014-08-26 so it should be fixed.
oVirt 3.5 has been released and should include the fix for this issue.