Deleted a node by mistake; what are my options now?

[Tony]

I renamed hostname on a proxmox node and then added it to a cluster. For some reason it showed in the cluster under the old name, so I deleted it from the cluster with pvecm delnode. It was a blunder from me, I didn't read the wiki carefully.

Now my question is: apart from reinstalling the node, is there any alternative? There is no VM on the node, but some ZFS pool with data.

 

[Andres.Urzagasti]

Hi!
The data in zfs pool, is from virtualized hosts? In this case you can not rejoining the cluster... node must be empty to join.
See this similar post: https://forum.proxmox.com/threads/c...s-host-already-contains-virtual-guests.55965/

 

[bbgeek17]

The "node must be empty" only refers to /etc/pve context. There is nothing fatal about having data on local storage.
It should be absolutely possible to rejoin the node. You should backup your data just in case.
PVE will not automatically delete/create/etc over your existing ZFS storage.
Look at the join procedures, read them completely before executing any steps. Try to understand why things are done and measure twice before cutting.

---

Blockbridge: Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

 

[Tony]

the wiki says:

If, for whatever reason, you want this server to join the same cluster again, you have to:

• do a fresh install of Proxmox VE on it,
• then join it, as explained in the previous section.

so I assume a fresh install is a must?

 

[fabian]

https://pve.proxmox.com/pve-docs/chapter-pvecm.html#pvecm_separate_node_without_reinstall if you follow the steps on the 'separated' node, you should be able to rejoin without issues.

 

[Tony]

@fabian thank you for your hint; I followed the steps but got stuck at this point:

pvecm add 192.168.11.12
Please enter superuser (root) password for '192.168.11.12': ************
detected the following error(s):
* this host already contains virtual guests
Check if node may join a cluster failed!

the content of /etc/pve/nodes is below (on all nodes):

ls /etc/pve/nodes
daika-pve  fenox-pve  maima-pve  teima  teima-pve

The problematic node is teima-pve. I renamed it from teima but I guess I should have rebooted after changing hostname, then join.

So I guess I need to:
- reboot the problematic node
- delete everything under /etc/pve/nodes except teima-pve
- delete teima and teima-pve from /etc/pve/nodes on a cluster node
- re-join teima-pve

Have I missed or overlooked anything?

 

[fabian]

something like that, yes. on the teima node you also need to remove any traces of clustering (according to the link I replied with earlier).

 

[Tony]

update: after following the above instructions I could rejoin the node without reinstalling. Thanks again for your help.