Actions
Bug #11907
closedcrushmap validation must not block the monitor
Status:
Closed
Priority:
High
Assignee:
-
Category:
Performance/Resource Usage
Target version:
-
% Done:
0%
Source:
other
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(RADOS):
Monitor
Pull request ID:
Crash signature (v1):
Crash signature (v2):
Description
Not only does it fail if the machine is slow, it will also systematically fail if the crushmap contains a ruleset that takes more than 5 sec to test no matter what: it can be reproduced by creating an erasure coded ruleset with a max_size > 20.
check_response: 96: '[' '' -a '!=' ']'
check_response: 101: grep --quiet -- 'Error EINVAL: Failed to parse crushmap: /tmp/cephtool12221/crushtool: timed out (5 sec)' /tmp/cephtool12221/test_invalid.12221
test_mon_crushmap_validation: 1581: ceph tell 'mon.*' injectargs --crushtool crushtool
*** DEVELOPER MODE: setting PATH, PYTHONPATH and LD_LIBRARY_PATH ***
make[2]: *** [check-recursive] Terminated
Updated by Loïc Dachary about 8 years ago
- Related to Bug #14136: "Error EINVAL: crushtool check failed with -22" in upgrade:infernalis-infernalis-distro-basic-openstack added
Updated by Greg Farnum almost 7 years ago
- Project changed from Ceph to RADOS
- Category set to Performance/Resource Usage
- Component(RADOS) Monitor added
Don't we internally time out crush map testing now? Does it behave sensibly if things take too long?
Actions