Proxmox VE cluster rejoin node

[g.migliaccio]

Hello there,
I made a mistake, I joined a single node to a two nodes cluster (this one is in production) but it was the wrong one (single node was without VM's) so I killed it without turning it off. Now the two nodes cluster seems to works fine, then i removed a VM from the custer and now i get an error message during the scheduled backup process ''Can't find the VMXXX''.
It's just an error message but the backup works. I still got the single node without vms and without any kind of mod since i killed it from the cluster, can I rejoin it to the cluster disable all the backup job isolate it from all the shared storage and then remove it again but in the correct way? Should this solve the issue?
Please can you help me to solve it?

INFO: Finished Backup of VM 114 (00:00:06)
INFO: Backup finished at 2025-01-26 22:49:01
ERROR: Backup of VM 119 failed - unable to find VM '119'
INFO: Failed at 2025-01-26 22:49:01
INFO: Backup job finished with errors
TASK ERROR: job errors

 

[cheiss]

Hi,

Now the two nodes cluster seems to works fine, then i removed a VM from the custer and now i get an error message during the scheduled backup process ''Can't find the VMXXX''.

First of, are you aware of the drawbacks of 2-node clusters and potentially add a QDevice?

Anyway, regarding the problem: Since you removed the VM, have you ticked the checkbox to purge it from job configurations when deleting?
You can also remove it yourself from the backup job, by editing /etc/pve/jobs.cfg and removing the ID of the removed VM from the vmid line of the appropriate backup job.

 

[g.migliaccio]

Thank you for the reply Cheiss, i deleted the vm119 form the "jobs.cfg" file.

First of, are you aware of the drawbacks of 2-node clusters and potentially add a QDevice?

This is the output for the command pvecm nodes:
root@dl380Gen10-1:~# pvecm nodes

root@dl380Gen10-1:~# pvecm nodes

Membership information
----------------------
    Nodeid      Votes    Qdevice Name
         1          1    A,V,NMW dl380Gen10-1 (local)
         2          1    A,V,NMW dl380gen10-2
         0          1            Qdevice
root@dl380Gen10-1:~#

Is everything correct? My target is to rebuild the one that i've detached with new settings like name and pve version (can i keep same ip address?) and then rejoin it to the 2nodes cluster (in production).
Thank for your time

 

[cheiss]

Is everything correct?

You have a QDevice configured, so yeah, that should be fine.
Just wanted to be sure

My target is to rebuild the one that i've detached with new settings like name and pve version (can i keep same ip address?)

Sure, using the same IP address should be no problem, at least if the node was properly removed from the cluster as described in our documentation.

 

[g.migliaccio]

Thank you! To be safe i will change also the ip address because I killed it without turning off so not pretty sure and I can't risk to compromise the production cluster