Project

General

Profile

Bug #38402

ceph-objectstore-tool on down osd w/ not enough in osds

Added by Sage Weil about 1 month ago. Updated about 1 month ago.

Status:
Verified
Priority:
High
Assignee:
-
Category:
-
Target version:
-
Start date:
02/20/2019
Due date:
% Done:

0%

Source:
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(RADOS):
Pull request ID:

Description

crush couldn't quite do it:

2019-02-19T22:55:52.500 INFO:tasks.thrashosds.thrasher:Recovered, killing an osd
2019-02-19T22:55:52.500 INFO:tasks.thrashosds.thrasher:Killing osd 7, live_osds are [1, 3, 2, 5, 4, 7, 6, 0]
2019-02-19T22:55:53.085 INFO:tasks.thrashosds.thrasher:Removing osd 7, in_osds are: [1, 3, 4, 7, 6]
2019-02-19T22:55:53.912 INFO:tasks.thrashosds.thrasher:Testing ceph-objectstore-tool on down osd
2019-02-19T22:55:55.981 INFO:tasks.thrashosds.thrasher:Waiting for clean again
...
2.1         247                  0      247         0       0 727153251  81       81 active+undersized+degraded 2019-02-19 22:55:53.885181 140'1381 151:3730   [4,6]          4   [4,6]              4   120'1082 2019-02-19 22:55:02.290721             0'0 2019-02-19 22:51:34.834103             0 
...
AssertionError: failed to become clean before timeout expired

/a/sage-2019-02-19_16:53:27-rados-wip-sage-testing-2019-02-19-0832-distro-basic-smithi/3613198

History

#1 Updated by Greg Farnum about 1 month ago

I see we have a bunch of live OSDs which are not marked in; perhaps this is the same as #37439?

Also available in: Atom PDF