[ceph-users] Re: Quincy full osd(s)

[Wesley Dillingham]

Can you send along the return of "ceph osd pool ls detail" and "ceph health
detail"

On Sun, Jul 24, 2022, 1:00 AM Nigel Williams 
wrote:

> With current 17.2.1 (cephadm) I am seeing an unusual HEALTH_ERR
> Adding files to a new empty cluster, replica 3 (crush is by host), OSDs
> became 95% full and reweighting them to any value does not cause backfill
> to start.
>
> If I reweight the three too full OSDs to 0.0 I get a large number of
> misplaced objects but no subsequent data movement, cluster remains at
> HEALTH_WARN "Low space hindering backfill". Cluster has 1200 OSDs (all
> except three are close to empty).
>
> Balancer is on, autoscale is on for pool.
>
> I feel I am overlooking something obvious, if anyone can suggest what it
> would be appreciated. thanks.
> ___
> ceph-users mailing list -- ceph-users@ceph.io
> To unsubscribe send an email to ceph-users-le...@ceph.io
>
___
ceph-users mailing list -- ceph-users@ceph.io
To unsubscribe send an email to ceph-users-le...@ceph.io

[ceph-users] Re: Quincy full osd(s)

[Nigel Williams]

Hi Wesley, thank you for the follow up.

Anthony D'Atri kindly helped me out with some guidance and advice and we
believe the problem is resolved now.

This was a brand new install of a Quincy cluster and I made the mistake of
presuming that autoscale would adjust the PGs as required, however it never
kicked into action. I then neglected to check the pools were configured
correctly and was surprised that after copying 50TB to the cluster it was
full. I went down the wrong rabbit-holes to diagnose, Anthony got me back
on track, thanks Anthony!

During this process Anthony also identified a documentation/setup issue and
will be doing some PRs to improve the documentation, the gist of it is that
the device-class needs to be set on the .mgr/.nfs pools in order for
autoscale-status to produce any output.
___
ceph-users mailing list -- ceph-users@ceph.io
To unsubscribe send an email to ceph-users-le...@ceph.io