Project

General

Profile

Actions

Bug #12737

closed

pg stuck in WaitLocalBackfillReserved, asok unresponsive

Added by Sage Weil over 8 years ago. Updated over 8 years ago.

Status:
Can't reproduce
Priority:
Urgent
Assignee:
-
Category:
-
Target version:
-
% Done:

0%

Source:
Q/A
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

several PGs were stuck in backfilling and wait_backfill state. They appeared blocked by osd.1, which had 2 osds in wait_backfill and WaitLocalBackfillReserved.

asok was unresponsive, but OSD appeared functional in all other respects.

strangely gdb ceph-osd $pid showed no threads... (trusty bare metal host)

'ceph osd down 1' made it come back to life, PGs peers, asok started working.

/a/sage-2015-08-19_19:13:53-rados-wip-sage-testing-distro-basic-multi/1023061

These PRs were in the mix at the time:

https://github.com/ceph/ceph/pull/5539
https://github.com/ceph/ceph/pull/5518
https://github.com/ceph/ceph/pull/5315
https://github.com/ceph/ceph/pull/5259
https://github.com/ceph/ceph/pull/3595

Actions #1

Updated by Sage Weil over 8 years ago

  • Status changed from Need More Info to Can't reproduce
Actions

Also available in: Atom PDF