Actions
Bug #9121
closedmds: inode stuck recovering after client restart
% Done:
0%
Source:
Development
Tags:
Backport:
Regression:
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(FS):
MDS
Labels (FS):
Pull request ID:
Crash signature (v1):
Crash signature (v2):
Description
2014-08-14 13:53:39.216244 7fd45d652700 10 mds.0.locker eval_gather (ifile excl->sync) on [inode 100021a930e [2,head] /teuthology-archive/sage-2014-08-13_21:04:38-rados-firefly-testing-basic-multi/423538/orig.config.yaml auth v16 ap=3+0 needsrecover recovering s=2402 n(v0 b2402 1=1+0) (ifile excl->sync) (iversion lock) cr={551831=0-4194304@1} caps={551831=p/-@4} | ptrwaiter=1 request=1 lock=1 caps=1 waiter=1 authpin=1
0x19091200]
replay at 2014-08-14 10:34:46.270573
logs on burnupi21:~ubuntu/9121/100021a930e (grepped)
Updated by Sage Weil over 9 years ago
- Status changed from New to In Progress
- Assignee set to Sage Weil
recovery is working.. there are just a lot of inodes queued:
2014-08-14 14:40:06.695087 7fd45f757700 10 mds.0.cache do_file_recover 128612 queued, 4 recovering
so it will take a while. we should bump the locks that need immediate attention to the top of the queue, or even force them to be processed immediately.
Updated by Sage Weil over 9 years ago
- Status changed from In Progress to Resolved
Actions