Bug #39546
closed
Warning about past_interval bounds on deleting pg
Added by Samuel Just about 5 years ago.
Updated over 4 years ago.
Description
cluster [ERR] 4.7
required past_interval bounds are empty [228,226) but past_intervals is not: ([166,225]
all_participants=1,3,5 intervals=([166,182] acting 1,3),([183,225] acting 3,5))
This pg took long enough to be deleted that the oldest stored map ended up past les. I think the simple fix is for check_past_interval_bounds to just not warn if deleting.
sjust-2019-04-26_14:00:33-rados-wip-sjust-peering-refactor-distro-basic-mira/3897200/
- Project changed from Ceph to RADOS
- Assignee deleted (
Samuel Just)
- Related to Bug #37509: require past_interval bounds mismatch due to osd oldest_map added
- Has duplicate Bug #41176: required past_interval bounds are empty .. but past_intervals is not added
/kchai-2019-08-12_14:45:48-rados-wip-kefu-testing-2019-08-12-1306-distro-basic-mira/4211285/
/a/sage-2019-08-19_13:35:06-rados-wip-sage-testing-2019-08-17-1023-distro-basic-smithi/4230273
- Pull request ID set to 29747
- Status changed from New to Pending Backport
- Backport set to mimic, nautilus
- Copied to Backport #41502: mimic: Warning about past_interval bounds on deleting pg added
- Copied to Backport #41503: nautilus: Warning about past_interval bounds on deleting pg added
- Status changed from Pending Backport to Resolved
While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are in status "Resolved" or "Rejected".
Also available in: Atom
PDF