Dear Ceph community,
We are having an issue with the MGR progress module:
Module 'progress' has failed: ('e7fb29e3-9caf-4b20-b930-cee8474526bb',)
We are currently on ceph version 16.2.7
(dd0603118f56ab514f133c8d2e3adfc983942503) pacific (stable).
I'm aware that there are already issues and a fix, which I assume
corresponds to our problem:
https://tracker.ceph.com/issues/54267
https://tracker.ceph.com/issues/53803
https://github.com/ceph/ceph/pull/44672
But I'm unsure how and when to apply this fix.
First, there hasn't been a pacific release since version 16.2.7 (which
we are already on).
So would I need to update using a git branch based image like this?
ceph orch upgrade start --image quay.io/ceph-ci/ceph:pacific
Second, do I actually need to upgrade right away?
Or can I somehow first reset the progress module or the events dictionary,
which leads to the race condition (missing event) mentioned in the fix.
In particular since it is highly recommended to upgrade in a healthy state,
I would be hoping to fix/reset this issue temporarily before applying
the upgrade.
I tried resetting progress as follows without success:
0|0[root@osd-1 ~]# ceph progress
Error EIO: Module 'progress' has experienced an error and cannot
handle commands: ('e7fb29e3-9caf-4b20-b930-cee8474526bb',)
0|5[root@osd-1 ~]# ceph progress clear
Error EIO: Module 'progress' has experienced an error and cannot
handle commands: ('e7fb29e3-9caf-4b20-b930-cee8474526bb',)
0|5[root@osd-1 ~]# ceph progress off
Error EIO: Module 'progress' has experienced an error and cannot
handle commands: ('e7fb29e3-9caf-4b20-b930-cee8474526bb',)
0|5[root@osd-1 ~]# ceph mgr module disable progress
Error EINVAL: module 'progress' cannot be disabled (always-on)
Is there any other way to get rid of this event?
Thank you very much for your input.
Best, Mathias
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
