Actions
Bug #4670
closedlibrbd: DiffIterateStress test failing
% Done:
0%
Source:
Q/A
Tags:
Backport:
Regression:
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):
Description
it only happens when you do the full test set and not when you run just that test because the random seed is different.
appears to be related to how the test is calculating the written vs exists sets. it writes and then zeros an object during the same interval and thinks the object should exist, but it does not (and list_snaps gets ENOENT).
Actions