Upgrade from 6.2 to 7.0 with new installation (CEPH monitor issue nautilus <> pacific)

[rcrack2k]

Hi guys,

i've replaced 3 nodes of my 6 node-cluster with fresh installs of Proxmox 7.0.

CEPH runs as version 14.2.22 on the v6.2 nodes
and on the v7.0 nodes CEPH runs as version 16.2.5.

Now, when i add a new CEPH-MONITOR on the new host (ceph v16.2.5), RAM - on the old machines (ceph 14.2.22) that are currently running the monitor deamons - will use up to 100% of 128 GB and the machine gets stale.

How can i move my monitors from the old CEPH to the new CEPH Hosts?

Will post some Screenshots and Info-Messages from MDS Daemon when i am getting access to the logs.



The MONITOR Process on HOST-3 hangs at

2021-09-26 19:28:47.165 7f294e3a4700  1 mon.host-03@1(probing) e6  adding peer [v2:192.168.110.4:3300/0,v1:192.168.110.4:6789/0] to list of hints

and will fill the logfile /var/log/ceph/ceph-mon.host-03.log endless!

 

[rcrack2k]

Now here are the PROXMOX Screenshots of CPU and RAM usage of the old host machines:



CEPH MON-Deamons on htop (you can see that CEPH-MONITOR is using all RAM):


Is it save to upgrade CEPH packages on my old PROXMOX Host-Machines?