Bug 1059904
Summary: | Inconsistent blockInfo values when the VM is finalizing live migration | ||
---|---|---|---|
Product: | Red Hat Enterprise Linux 6 | Reporter: | Chris Pelland <cpelland> |
Component: | libvirt | Assignee: | John Ferlan <jferlan> |
Status: | CLOSED ERRATA | QA Contact: | Virtualization Bugs <virt-bugs> |
Severity: | high | Docs Contact: | |
Priority: | high | ||
Version: | 6.5 | CC: | abaron, acathrow, amureini, bazulay, bili, cpelland, dallan, djasa, dyuan, eblake, fsimonce, iheim, jferlan, jiahu, lcui, lpeer, michal.skrivanek, mjenner, mzhan, pm-eus, scohen, tlavigne, ydu, yeylon, zpeng |
Target Milestone: | rc | Keywords: | ZStream |
Target Release: | --- | ||
Hardware: | Unspecified | ||
OS: | Unspecified | ||
Whiteboard: | |||
Fixed In Version: | libvirt-0.10.2-29.el6_5.4 | Doc Type: | Bug Fix |
Doc Text: |
Cause: Migration of transient guests returned the default "allocation" virDomainBlockInfo value for a qcow2 block backed device instead of the actual from a virDomainGetBlockInfo() call in a brief instance between when libvirt declared the guest migrated and performing the guest removal.
Consequence: Applications that depended upon a non default value being returned may take the wrong action, such as extending the logical volume since the allocation would be the same as physical.
Fix: When the condition is detected for a transient domain return a failure.
Result: The virDomainGetBlockInfo() will fail with the error "domain is not running" for a transient domain once the migration has successfully completed and before the domain is destroyed on the source.
|
Story Points: | --- |
Clone Of: | Environment: | ||
Last Closed: | 2014-02-20 09:50:23 UTC | Type: | --- |
Regression: | --- | Mount Type: | --- |
Documentation: | --- | CRM: | |
Verified Versions: | Category: | --- | |
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
Cloudforms Team: | --- | Target Upstream Version: | |
Embargoed: | |||
Bug Depends On: | 1040507, 1065531 | ||
Bug Blocks: | 871829, 1038069 |
Description
Chris Pelland
2014-01-30 22:09:42 UTC
Hi Federico, We are trying to verify this bug, but we can not reproduce the bug under pure libvirt environment. So can you help to have a try with the package libvirt-0.10.2-29.el6_5.4 in rhevm? I think the expected result should fail to get domblkinfo with error info including "domain is not running". Thanks in advance. EircLee The libvirt-0.10.2-29.el6_5.4 build fixes the original issue: doesn't occur anymore on environment where it was spotted first when libvirt in both hosts is updated to this version. (In reply to David Jaša from comment #9) > The libvirt-0.10.2-29.el6_5.4 build fixes the original issue: doesn't occur > anymore on environment where it was spotted first when libvirt in both hosts > is updated to this version. Thanks for the test. John, Do you think that enough to verify this bug? Or we need catch the error info:"domain is not running"? Thanks, EricLee Since the problem was seen in the rhev test environment primarily and they indicate the patch works, I think that's good enough. The error message about domain is not running would be somewhat generic - you may not know that it came from this path. Returning the error though results in their thread no fetching data which ultimately is what led to the issue. If you're more curious about the core issue see the bz this was copied from (bz 1040507). Thanks John. Setting VERIFIED as comment #9 and comment #11. Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHBA-2014-0193.html |