On 7/24/23 19:17, Humberto Jose de Sousa wrote:
Hi Aaron,

It was OSDs Bluestore.
I tried only `ceph-volume simple scan /dev/sdx` and it didn't work.

"simple" is for legacy file store OSDs :)

In the end I decided to destroy and recreate.

This problem only occurred with one of the five updates. I didn't
understand ...

Thanks!
Em seg., 24 de jul. de 2023 às 05:58, Aaron Lauterer <[email protected]>
escreveu:

Are those OSDs bluestore ones or still old file-store ones?

Have you tried running `ceph-volume lvm activate --all`?

It should search for bluestore OSDs and activate them. Activating means,
mounting the tmpfs into the /var/lib/ceph/osd/ceph-{id} and starting the
systemd
units.

On 7/21/23 21:56, Humberto Jose de Sousa wrote:
Hi folks,

I upgraded my ceph cluster from Octopus to Pacific and Pacific to Quincy,
no problems here.

So I upgraded pve 7 to 8. After reboot the OSDs don't up. There is no
directory /etc/ceph/osd/ and json files. The directory
/var/lib/ceph/osd/ceph-<id>/ is empty.

How can I recreate the lost files?





_______________________________________________
pve-user mailing list
[email protected]
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-user

Reply via email to