Question on clusters and mixing AMD/Intel

[NewDude]

I'm migrating a cluster from XCP-ng to Proxmox. I'm using Vinchin for this and for the most part it's not too bad, provided I don't need to boot from a Centos CD and reconfigure the start environment to include VirtIO drivers - I've done this twice and am working to not have that happen a third time.

Vinchin works, but is quirky. So far I've had a couple of VMs that wouldn't restore to the new Proxmox cluster node (Intel) where others did so just fine, but they would restore to an AMD box I also have Proxmox in.

So now I've got VMs on two incompatible Proxmox hosts, backing up happily to pbe, and I'm looking for the best way to migrate those to the failover cluster that will be running once I've gotten everything off XCP-ng and converted all the nodes to Proxmox.

I'm here because by default I'd just shut down the VM, back up to pbe, then restore to the cluster and call it done. But Grok is insisting I can join an AMD server with existing VMs to a cluster, but I need to do it as a non-HA node.

I think it's hallucinating, but if I could do this long-term management would be easier, as would the migration.

Is it possible to join an AMD server with VMs to a cluster and keep those VMs, provided I don't turn on HA?

 

[leesteken]

It in the manual: https://pve.proxmox.com/pve-docs/pve-admin-guide.html#pvecm_join_node_to_cluster

All existing configuration in /etc/pve is overwritten when joining a cluster. In particular, a joining node cannot hold any guests, since guest IDs could otherwise conflict, and the node will inherit the cluster’s storage configuration.

 

[NewDude]

That confirms my understanding. Thanks!