Ceph v13.2.4 issue with snaptrim

[Vytautas Jonaitis <vytautas.j@xxxxxxxxxxx>]

Hello,

 

Few month ago we experienced issue with Ceph  v13.2.4:

 

1. One of the nodes had all it's osd's set to out. To clean them up for replacement.
2. Noticed that a lot of snaptrim was running.
3. Set nosnaptrim flag on the cluster (to improve performance).
4. Once mon_osd_snap_trim_queue_warn_on appeared, removed nosnaptrim flag.
5. All osds on the cluster crashed and started flapping. Set nosnaptrim flag back on.

 

Issue it registered in tracker, additional logs were collected - https://tracker.ceph.com/issues/38124

However, it is still present.

 

What options do I have? I would like to know when/if issue would be fixed (it was not in v13.2.5 release), or, alternatively, to contact developer who can resolve it.

 

--

Best regards,

Vytautas J.

 

_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com