Bug #24222: Manager daemon y is unresponsive during teuthology cluster teardown - RADOS - Ceph

Actions

Copy link

Bug #24222

closed

Manager daemon y is unresponsive during teuthology cluster teardown

Added by Sage Weil almost 6 years ago. Updated almost 6 years ago.

Status:

Resolved

Priority:

High

Assignee:

Category:

Target version:

% Done:

Source:

Tags:

Backport:

mimic,luminous

Regression:

Severity:

3 - minor

Reviewed:

Affected Versions:

ceph-qa-suite:

Component(RADOS):

Pull request ID:

Crash signature (v1):

Crash signature (v2):

Description

2018-05-22 00:21:18.397796 mon.b mon.1 172.21.15.70:6789/0 121 : cluster [WRN] Manager daemon y is unresponsive.  No standby daemons available.
2018-05-22 00:21:18.451864 mon.b mon.1 172.21.15.70:6789/0 122 : cluster [DBG] mgrmap e4: no daemons active

at very end of ceph.log. meanwhile, teuthology.log,

2018-05-22T00:19:38.372 INFO:teuthology.misc:Shutting down mds daemons...
2018-05-22T00:19:38.373 INFO:teuthology.misc:Shutting down osd daemons...
...
2018-05-22T00:20:33.291 INFO:teuthology.misc:Shutting down mgr daemons...
...
2018-05-22T00:20:33.465 INFO:teuthology.misc:Shutting down mon daemons...
...
2018-05-22T00:21:52.789 INFO:tasks.ceph:Checking cluster log for badness...

hmm, just before shutdown, we have

2018-05-22T00:19:36.598 INFO:teuthology.orchestra.run.smithi050:Running: "sudo ceph --cluster ceph tell 'mon.*' injectargs -- --no-mon-health-to-clog"

maybe this needs to include mgr and other daemon health?

/a/sage-2018-05-21_18:33:19-rados-wip-sage3-testing-2018-05-21-1130-distro-basic-smithi/2563080

Related issues 2 (0 open — 2 closed)