These hosts have new drives, let's learn what we can about whether they are now better or worse.
Description
Description
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | dcaro | T334240 [cloudceph] Slow operations - tracking task | |||
| Resolved | taavi | T348634 ceph slow ops 2023-10-11 | |||
| Unknown Object (Task) | |||||
| Unknown Object (Task) | |||||
| Restricted Task | |||||
| Resolved | dcaro | T386725 evaluate new drives in cloudcephosd102[123] |
Event Timeline
Comment Actions
I did a first check of the current values for the smartcl reported counters, all look good so far (no more Offline_Uncorrectable_Errors), there's one drive with some Command_Timeout, but it's very low, so no worries.
I'm going to try to have a look at some performance metrics, to see if at least there's no regression there either.
Comment Actions
Just checked the number of operations/s (as a proxy for performance):
- For cloudcephosd1021, comparing with 1018, there's a bit of an improvement:
- For cloudcephosd1022, comparing (more or less) with 1018, things seem to be like before (no improvement, but not worse):
- For cloudcephosd1023, compared (more or less) with 1020, more or less the same (difference gravitating towards 0):
So I think it's ok to keep these in prod :)