How to safely reboot a node with the ceph cluster being still available

[codingspiderfox]

I installed 3 proxmox hosts with ceph 16.2 and mons on pmxhost1 pmxhost2, mgr on pmxhost1 and pmxhost2
Each node has 2 equally sized OSDs

I set the 2 OSDs to "out" on the node before reboot. Then I shut down the pmxhost2.
When opening the Proxmox UI on pmxhost1, all the ceph dasboard on pmxnode1 gives me the error "got timeout (500)".

What am I doing wrong?

 

[gurubert]

You need another MON, otherwise the remaining one is not able to form a quorum and will refuse to work.

 

[codingspiderfox]

thx, will try that