Hi all.
I'm running version 14.2.16.
My cluster is remapping and backfilling after "reweight-by-utilization".
This morning I started to have some issues:
- all my mgr active and standby crashed
- MDS stopped reporting statistics
- MON is using lot of memory
- Cluster was very slow
I restarted everything but osd
I managed to get the statistics back, restarting the MDS a couple of times.
All above issues are gone but I have this "mon.node1 (mon.0) 1529395 :
cluster [DBG] osdmap e??????:" about 1 every 2 second with a different
epoc.
Question:
Is it normal or there is something wrong?
Should be the osdmap more "stable" ?
What can change the osdmap?
Thanks for the help
Andrea
Show replies by date