Project

General

Profile

Bug #6593

osd: copy-from object blocking clashes with recovery

Added by Greg Farnum almost 6 years ago. Updated almost 6 years ago.

Status:
Duplicate
Priority:
High
Assignee:
-
Category:
-
Target version:
-
Start date:
10/18/2013
Due date:
% Done:

0%

Source:
Development
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:

Description

$ ceph -w
    cluster 9a2d8067-5c2d-4274-98a4-45c4d731067a
     health HEALTH_WARN 1 pgs recovering; 1 pgs stuck unclean; 8 requests are blocked > 32 sec; recovery 11/33276 objects degraded (0.033%); 2/16638 unfound (0.012%); pool data pg_num 164 > pgp_num 34; pool metadata pg_num 64 > pgp_num 44; pool rbd pg_num 84 > pgp_num 64; pool radosmodel-0 pg_num 51 > pgp_num 41
     monmap e1: 3 mons at {a=10.214.131.31:6789/0,b=10.214.131.28:6789/0,c=10.214.131.31:6790/0}, election epoch 6, quorum 0,1,2 a,b,c
     mdsmap e5: 1/1/1 up {0=a=up:active}
     osdmap e927: 6 osds: 6 up, 6 in
      pgmap v1332: 363 pgs, 4 pools, 28678 MB data, 16638 objects
            3835 MB used, 2606 GB / 2749 GB avail
            11/33276 objects degraded (0.033%); 2/16638 unfound (0.012%)
                 362 active+clean
                   1 active+recovering

2013-10-18 13:45:38.635777 osd.3 [WRN] slow request 960.971893 seconds old, received at 2013-10-18 13:29:37.663825: osd_op(osd.3.0:1 plana0925762-273 [assert-version v1151,copy-get max 8388608 ver 0] 3.69f392f2 e729) v4 currently waiting for missing object

(That one pg has been blocked for >10 minutes.)

Saw this when trying to reproduce #6391. I didn't have logging enabled, but I assume that there's a conflict between the object being blocked and recovery in some way or other.


Related issues

Related to Ceph - Bug #6585: osd: backfill vs copy-from delay badness (was osd: ENOENT on clone) Resolved 10/17/2013

History

#1 Updated by Sage Weil almost 6 years ago

  • Status changed from New to Duplicate

sounds like a dup of #6585

Also available in: Atom PDF