Replacing Main Cluster Node

cocobanana

Member
Jun 18, 2021
29
3
8
37
Hi,

I have questions here. Apparently I have 3 Nodes with cluster

- Node-01
- Node-02
- Node-03

I create cluster name at the Node-01.

Node-02 and Node-03 Join Cluster to Node-01 so the keyprint is based on the Node-01

Now I am having super problem which means I need to remove the Physical Server ( Node-01 ) and replace it to new one.

May I know how can i accomplish that very safe without breaking any cluster or without bring any problems for the cluster , keyprint and e.t.c

Please help


** I am aware that I need to make sure that there is no VM running in Node-01.
 
Please follow the steps detailed in the docs: https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_remove_a_cluster_node
If you don't plan on reinstalling the node, then follow the steps in section 5.5.1 Separate a Node without Reinstalling.

Hi,
In my case, how about if I replace the Node-01 ( Node-02 and Node-03 join the cluster using the Node-01 information ) to a new server, is it possible? And can I used the Steps that you provide? How about the Joint Information and Fingerprint?
 
You can use any node in the cluster. There's no `main` node in PVE.
 
You can use any node in the cluster. There's no `main` node in PVE.
Hi,
Means If I remove the Node-01 and install proxmox to a fresh and new server and using Node-01 as a computer name and join the cluster at Node-02 or Node-03, is it possible?
 
Yes, that's possible if you follow the steps detailed in the docs.
 
  • Like
Reactions: cocobanana
Yes, that's possible if you follow the steps detailed in the docs.
Hi,

If I plan to use the same name ( Node-01 ) with a fresh and new server and higher spec, should i follow

5.5. Remove a Cluster Node or 5.5.1. Separate a Node Without Reinstalling ?

Please help​

 
I'll quote the docs:
As mentioned above, it is critical to power off the node before removal, and make sure that it will not power on again (in the existing cluster network) with its current configuration. If you power on the node as it is, the cluster could end up broken, and it could be difficult to restore it to a functioning state.
So if you don't plan on reusing the node, e.g. by simply readding it afterwards, then follow the steps described in 5.5.