My next idea is using a 1/6 cluster, then tainted and drain the old agents - then remove the old agents so i'am back to 1/3 and can decommission the old agents. Now comes the tricky part i'am not sure about. I would somehow "clone" the disk or preferable the state of server1 to the new clusterB-server1 (maybe just the var/lib/rancher) folder, then bring it back up there and somehow change the IP from 10.10.5.50 to 10.3.1.100 (etcd updated) The latter is my actual question/issue, can i do this/how would i go about it?

If interesting, it, it's all RKE2 1.33.6 with longhorn as CSI and calico as CNI

i could also use velero to backup the old server and restore it on the new one i assume, that's the other option.

So can i a) backup the state of clusterA server1 (10.10.5.50) and restore it on clusterB server2 (10.3.1.100) and somehow adjust the ip (using velero or similar) ? b) i assume then change the "bootstrap server" in the config.yaml of each worker so they register / update their registration to 10.3.1.100 - is that correct?

Glad for any hints!

For 1) i could also use https://docs.rke2.io/upgrades/roll-back / rolling back using cluster reset on the new cluster node (the worker wont need it) - or use https://docs.rke2.io/datastore/backup_restore instead of velero (since it might do more?)

I see, i guess for my server 2 server migration there is an option https://docs.rke2.io/datastore/backup_restore#restoring-to-new-hosts - that solves that