Cannot start VM when cluster not ready

[neilsonwheels]

I think I've done something rather stupid!
I started my PVE journey with one node (called PVE1). I installed OPNsense as a VM and a few other things . . . all good.
I recently setup a couple of other nodes and made a cluster . . . again all good.
I had to shutdown the cluster.
When I powered on PVE1, it would not start the OPNsense VM because PVE1 could not communicate with the other nodes in the cluster.
I tried starting the VM from the command line and got a message saying "cluster not ready - no quorum".
None of the network interfaces are working on the host (because OPNsense is not running).
I there something I can do from the command line?
Can I leave the cluster? If so how? I'm thinking that if I do that, it will allow me to start the OPNsense VM.
I'm not too concerned about the other nodes, I just need OPNsense running.
I have backups, so worst case I guess I can wipe PVE1 and start over.

 

[gurubert]

AFAIK you can only remove nodes from a working cluster: https://pve.proxmox.com/wiki/Cluster_Manager#_remove_a_cluster_node

You may have luck following the chapter "Separate a Node Without Reinstalling".

 

[neilsonwheels]

For anyone following along at home, I managed to resolve the problem by issuing the following command:

pvecm expect 1

Immediately after doing that, all VMs that were set to start on boot started, including the OPNsense VM which then enabled all network interfaces on the PVE host.

Does anyone know how to make this a permanent change via a config?

 

[gurubert]

You should remove the corosync cluster configuration from that host.