I'm frustrated by health output like this:
id: 98463a16-36f5-4a1e-a232-64f4e987b5a2
health: HEALTH_WARN
16671/54990 objects misplaced (30.316%)
Degraded data redundancy: 8 pgs unclean
...
pgs: 16671/54990 objects misplaced (30.316%)
6 active+remapped+backfill_wait
2 active+remapped+backfilling
The PGs are 'unclean' because they don't have the 'clean' flag. But they
aren't degraded; just misplaced, and we already report that in the
misplaced health message.
I tried changing it to a different health message:
health: HEALTH_WARN
14948/54990 objects misplaced (27.183%)
Unclean state: 7 pgs unclean
...
pgs: 14948/54990 objects misplaced (27.183%)
6 active+remapped+backfill_wait
1 active+remapped+backfilling
1 active+clean
...but that looks like it'd be just as confusing to an administrator.
(What does "unclean" mean? What do I do about it?)
Perhaps we should just ignore the 'clean' state? I'm having a hard time
coming up with a reason the PG would be unclean that isn't reflected by
another message. I think the only example I can come up with is the
permanently 'remapped' state.. but even there we set 'clean'.
Can we just drop it?
https://github.com/ceph/ceph/pull/18493
sage
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
