Proxmox 4.4 Cluster Redo

cwoelkers

Active Member
Aug 17, 2017
19
2
43
Here is what I am trying to do, and how I seem to have failed.
I manage 4 hypervisors. One is brand new and 4.4 was installed on it. The other 3 are all older, 2 with 3.2, and one with 4.1. The plan was to move the VMs, a mix of containers and qemu, from one of the old hypervisors to the new one via backup and restore, refresh the old one with the latest OS and Proxmox 4.4, than move the VMs back to the original server and repeat with the others till all are running 4.4.
Once that was done I was going to cluster them all together for ease of management if nothing else.

So of course instead of doing all that moving I decided to cluster the first two once the OS and Proxmox was verified as working on the older one. This way, I thought, I could migrate the VMs instead of moving them via backup and restore. Unfortunately this is where I made my mistake. I wanted to make the cluster ring use our management VLAN but when the cluster was created it went to our main server VLAN instead.
Looking into moving this around, pve dot proxmox dot com/wiki/Separate_Cluster_Network, I discovered that /etc/pve was read-only, even for root.
So I looked into recreating the cluster entirely, found out how via blog dot sjas dot de/posts/proxmox-delete-and-recreate-cluster.html, but that requires restarting the server which will bring our VMs down, most of which are production machines.
So I thought, lets move the VMs back to the original, now updated, hypervisor. But I apparently cannot make a backup of any VM because /etc/pve is still read-only.

So I ask the community, is there a way to get out of this without a reboot? Or am I stuck with my issue and no way out except a reboot, or worse.
Edit: Links reworked to get around new post no URL issue.
 
So I found out how to get /etc/pve to be read-write again on the new hypervisor. Found an older post dealing with a similar issue, https://forum.proxmox.com/threads/cluster-and-quorum.7786/#post-59362, and the fix is to set the expected votes for the cluster to 1. Doing so allows a quorum and you can then take care of anything that needs to be done. I assume that this will have to be set back later on, but perhaps not.
This may help to change the cluster network as well, albeit with a reboot. Any ideas on a no-reboot scenario? Specifically looking for a way to get out of rebooting my VMs.
 
Last edited:

About

The Proxmox community has been around for many years and offers help and support for Proxmox VE, Proxmox Backup Server, and Proxmox Mail Gateway.
We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get yours easily in our online shop.

Buy now!