Bug #63514
open
mds: avoid sending inode/stray counters as part of health warning for standby-replay
Added by Venky Shankar 7 months ago.
Updated 23 days ago.
Category:
Administration/Usability
Backport:
quincy,reef,squid
Description
Otherwise a health warning from a s-r daemon looks like
sh-4.4$ ceph health detail
HEALTH_WARN 1 MDSs report oversized cache
[WRN] MDS_CACHE_OVERSIZED: 1 MDSs report oversized cache
cephfs(mds.0): MDS cache is too large (6GB/4GB); 0 inodes in use by clients, 0 stray files
sh-4.4$
Which is a wee bit confusing - so, avoid sending those counters as part of the warning.
Related issues
3 (3 open — 0 closed)
- Assignee set to Rishabh Dave
- Priority changed from Normal to High
Rishabh, you mentioned ins standup that you were looking at some trackers that need work on. Could you plan to priritize this next?
- Status changed from New to Fix Under Review
- Backport changed from quincy,reef to quincy,reef,squid
- Pull request ID set to 56065
- Status changed from Fix Under Review to Pending Backport
Venky, should we skip backporting this from Quincy? Or is it still valid?
Rishabh Dave wrote in #note-6:
Venky, should we skip backporting this from Quincy? Or is it still valid?
Let's backport (we can always close the backport PR if/when it reached EOL)
- Copied to Backport #65919: quincy: mds: avoid sending inode/stray counters as part of health warning for standby-replay added
- Copied to Backport #65920: squid: mds: avoid sending inode/stray counters as part of health warning for standby-replay added
- Copied to Backport #65921: reef: mds: avoid sending inode/stray counters as part of health warning for standby-replay added
- Tags set to backport_processed
Also available in: Atom
PDF