For some reasons, when I first installed my 3-node cluster at Hetzner, the management IP's were assigned directly on the interface instead of on a bridge. I guess that’s because I started with Debian and converted that to Proxmox later on and have missed that.
Now I have changed that on just one node (move the IP to a bridge) but if I login on that node (which works just fine!) I get a No route to host (595) accessing the other nodes. If I login to the other 2 nodes, I get the same error accessing the changed node.
Don’t get this error, as the corosync is a on completely different interface and seems to work just fine. Ceph is OK too. Also, I can access all three management interface IP’s on port 8006 and SSH with no issues at all.
What am I overlooking here?
Now I have changed that on just one node (move the IP to a bridge) but if I login on that node (which works just fine!) I get a No route to host (595) accessing the other nodes. If I login to the other 2 nodes, I get the same error accessing the changed node.
Don’t get this error, as the corosync is a on completely different interface and seems to work just fine. Ceph is OK too. Also, I can access all three management interface IP’s on port 8006 and SSH with no issues at all.
What am I overlooking here?
Last edited: