Project

General

Profile

Actions
Copy link

Bug #15590

closed

bluestore: shards missing after a few hours of testing

Added by Dan van der Ster about 8 years ago. Updated about 7 years ago.

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

0%

Source:
Community (user)
Tags:
bluestore
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
v10.2.0
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

One out our 18 servers is configured with 30x bluestore OSDs. (ceph-disk prepare --bluestore ...)

After a few hours of burn in testing with out rados striper client we get tons of shards missing on an 8+3 EC pool, like this:

2016-04-25 15:06:53.730627 osd.335 128.142.25.113:6915/379125 36 : cluster [ERR] 17.396s0 shard 128(2) missing 17:69c09e26::::head

ceph.log is here: ceph-post-file: ec2e97a8-c5fb-4e08-bd0b-6f33bbc0fbb0

There are no kernel disk errors, and no errors in any of the bluestore OSD logs.

I'm totally new to bluestore so would appreciate any help to debug this.

Actions
Copy link
 #1

Updated by Dan van der Ster about 8 years ago

In case it helps, here's the output of ceph-bluefs-tool on osd.128:

ceph-post-file: d3b2843e-e275-48d7-bc22-957751c4f80e

Actions
Copy link
 #2

Updated by Sage Weil about 7 years ago

  • Status changed from New to Can't reproduce
Actions
Copy link

Also available in: Atom PDF