Can't join a cluster

[Yvon]

Hello,

I was running a 3 nodes cluster perfectly fine then this week end one of my nodes couldn't boot. From what I've seen, it was missing a system file and was stuck. I installed PVE again and upgraded it to 5.3-8 and I want to join the cluster again but it seems I can't send the infos :

The "Join Information" is grayed and I don't understand why.

If I want to add PVE2 back again should I need to exclude it from the cluster first ?


PS : I upgraded the other 2 nodes to the same PVE version.

 

[Chris]

Hi,
you should first remove the failed (old) node from the cluster. Please follow the instructions here https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_remove_a_cluster_node. Then you can add the reinstalled (new) node to the cluster as described here https://pve.proxmox.com/pve-docs/pve-admin-guide.html#pvecm_join_node_to_cluster

 

[Chris]

Also note: The join information is only visible on the nodes of the cluster you want to join, not the new node which is not part of the cluster yet.

 

[Yvon]

Ok I found how to do it :

First remove the faulty node from the cluster : pvecm delnode pve2
Then add it back from the GUI with "Join Information" .

Well now it's called "pve" instead of "pve2" but it's working :

 

[Yvon]

Thank you Chris, it's working perfectly fine now.

 

[Chris]

You're welcome.

 

[okieunix1957]

This does not work for me, I still have a node that shows up in the cluster and the join info is blanked out when I tried to delnode it cannot find it. The server is completely powered off.

 

[Chris]

This does not work for me, I still have a node that shows up in the cluster and the join info is blanked out when I tried to delnode it cannot find it. The server is completely powered off.

Hi,
please check the /etc/pve/corosync.conf. Is the node still listed? Also check if the node is listed under /etc/pve/nodes.

 

[okieunix1957]

Hi,
please check the /etc/pve/corosync.conf. Is the node still listed? Also check if the node is listed under /etc/pve/nodes.

This issue is resolved. turns out the cluster name got changed.
Thanks for responding.