Actions
Bug #15938
closedrbd_mirror_stress: possible inconsistent replay after client crash
Status:
Resolved
Priority:
Urgent
Assignee:
Jason Dillaman
Target version:
-
% Done:
0%
Source:
other
Tags:
Backport:
jewel
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):
Description
During a write-after-write operation, the librbd cache writeback handler does not properly wait for the overwrite journal event to be safely committed to disk before flagging the overwritten event as complete (due to the overwrite). As a result, if a client (e.g. rbd bench-write) is killed after an overwrite of an event but before the new event is safely committed, the image will be inconsistent with its replicated version.
Updated by Jason Dillaman almost 8 years ago
- Status changed from In Progress to Fix Under Review
- Backport set to jewel
Updated by Mykola Golub almost 8 years ago
- Status changed from Fix Under Review to Pending Backport
Updated by Nathan Cutler almost 8 years ago
- Copied to Backport #15955: jewel: rbd_mirror_stress: possible inconsistent replay after client crash added
Updated by Jason Dillaman almost 8 years ago
- Status changed from Pending Backport to Resolved
Actions