Ceph

ushiromawashi

New Member
Mar 12, 2025
1
0
1
Hi,

I have some problems with Ceph mon.
1741770823493.png

How it happens?
Yesterday I want create osd disk on cluster nodes, but after created on the one node a got all time timeout in ceph. I search it and find that version 18 have some problems.
So I want to back to 17, and uninstall all ceph and conf :(, now after new install ceph cant start and have many problems.


Someone can help? Or must I reinstall all 6 nodes?
 
Welcome to the Proxmox forum, ushiromawashi!

Which problems did you encounter with Ceph Reef exactly? Do you have the Ceph log of the previous installation available? If so, please post it.

If I understood it correctly, downgrading Ceph and using the disks is not really possible and not supported. So, if you want to go with Ceph 17, you'd have to reinstall and recreate the OSDs, which also would mean that the data on them would need to be migrated away and then back on the OSDs again, if there was any data on them already.

Keep in mind that Ceph Quincy (version 17) is not maintained anymore as the end of life was on January 13th 2025, so it'd be good advise to use either Ceph Reef (version 18) or the newer Ceph Squid (version 19).
 
  • Like
Reactions: UdoB