Ceph

02/17/2022

09:33 AM RADOS Bug #54172: ceph version 16.2.7 PG scrubs not progressing

I would just like to add that scrubs started all of the sudden and the cluster is HEALTH_OK again. Daan van Gorkum

02/11/2022

12:18 AM RADOS Bug #54172: ceph version 16.2.7 PG scrubs not progressing

Added the logs for OSD 12,23,24 part of pg 4.6b. I don't think the logs are from the beginning when the osd booted an... Daan van Gorkum

02/08/2022

03:08 AM RADOS Bug #54172: ceph version 16.2.7 PG scrubs not progressing

Some more information on the non-default config settings (excluding MGR):... Daan van Gorkum

03:02 AM RADOS Bug #54172: ceph version 16.2.7 PG scrubs not progressing

One more thing to add is that when I set the noscrub, nodeep-scrub flags the pgs actually don't stop scrubbing either... Daan van Gorkum

12:16 AM RADOS Bug #54172: ceph version 16.2.7 PG scrubs not progressing

Correction. osd_recovery_sleep_hdd was set to 0.0 from the original 0.1. osd_scrub_sleep has been untouched. Daan van Gorkum

12:12 AM RADOS Bug #54172: ceph version 16.2.7 PG scrubs not progressing

This is set to 0.0. I think we did this to speed up recovery after we did some CRUSH tuning.... Daan van Gorkum

02/07/2022

02:46 AM RADOS Bug #54172: ceph version 16.2.7 PG scrubs not progressing

Additional information:
I tried disabling all client I/O and after that there's zero I/O on the devices hosting th... Daan van Gorkum

02:45 AM RADOS Bug #54172 (Resolved): ceph version 16.2.7 PG scrubs not progressing

A week ago I've upgraded a 16.2.4 cluster (3 nodes, 33 osds) to 16.2.7 using cephadm and since then we're experiencin... Daan van Gorkum