Bug #38783
Changing mon_pg_warn_max_object_skew has no effect.
% Done:
0%
Source:
Community (user)
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(RADOS):
Pull request ID:
Crash signature (v1):
Crash signature (v2):
Description
[root@ceph-mon1001 ~]# ceph health detail HEALTH_WARN 1 pools have many more objects per pg than average MANY_OBJECTS_PER_PG 1 pools have many more objects per pg than average pool images.rgw.buckets.data objects per pg (97733) is more than 12.5621 times cluster average (7780)
Tried to get rid of it by changing mon_pg_warn_max_object_skew setting to -1, 0 or 100000 but warning still in place.
[root@ceph-mon1001 ~]# ceph --admin-daemon=/var/run/ceph/cluster1-mon.ceph-mon1001.asok config get mon_pg_warn_max_object_skew { "mon_pg_warn_max_object_skew": "0.000000" }
Tried via config file, ejecting with ceph tell mon.*, restarted mons, nothing helped.
History
#1 Updated by xie xingguo almost 5 years ago
It's an mgr option. You should instead inject it to the mgr daemon.
#2 Updated by xie xingguo almost 5 years ago
- Status changed from New to 4
#3 Updated by Andrew Mitroshin almost 5 years ago
Injecting into mgr has solved the issue, thanks!
#4 Updated by Patrick Donnelly over 4 years ago
- Status changed from 4 to New
#5 Updated by Scott Hubbard over 3 years ago
Andrew Mitroshin wrote:
Injecting into mgr has solved the issue, thanks!
What command did you use to inject into the mgr? The command errors out for me.
$ sudo ceph tell mgr.* injectargs '--mon_pg_warn_max_object_skew 0'
no valid command found; 10 closest matches:
orchestrator service ls {<host>} {mon|mgr|osd|mds|iscsi|nfs|rgw|rbd-mirror} {<svc_id>} {json|plain} {--refresh}
orchestrator rgw rm <svc_id>
orchestrator osd create {<svc_arg>}
orchestrator osd rm <svc_id> [<svc_id>...]
insights prune-health <hours>
iostat
influx send
insights
influx config-set <key> <value>
influx config-show
Error EINVAL: invalid command
mgr.ceph-mon2: invalid command
#6 Updated by Sage Weil almost 3 years ago
- Project changed from Ceph to RADOS
- Category deleted (
Monitor)