Rename LVM storage in cluster (for live migration)

benutne

Active Member
Oct 5, 2018
6
0
41
47
I was getting bored with my current cluster which relies on an iSCSI SAN for shared storage. So I took a couple older R210ii Dell machines with a 1TB SSD, 1TB, HDD, and a small 200 GB OS drive and make a "cluster" out of them both. Except I cannot migrate anything between them because for some stupid reason I made their drives as LVM with their own names. Two boxes, identical except their hostname and their storage and no reason to even consider them a "cluster"

In each node I have:

pve1-HDD (pve1) - This is set as EXT4 and Directory as main storage for backups, ISOs, templates, etc.
pve1-SSD (pve1) - This is set as LVM/LVM-Thin as main storage for VMs

pve2-HDD (pve2) - This is set as EXT4 and Directory as main storage for backups, ISOs, templates, etc.
pve2-SSD (pve2) - This is set as LVM/LVM-Thin as main storage for VMs

I bet you can tell where this is going. So when I try and migrate from one to the other, it tells me the storage on the target location doesn't exist. It'd be just as easy to wipe and reinstall with the right names but I wanna figure this out the hard way.

Thoughts? And thanks in advance.
 
The easiest option is to backup any VMs that matter from one of the nodes, then delete all of that node's VMs. Then you should be able to click in the UI on Datacenter and go to the storage section. Remove the storage for that node and create it again with a name that matches the other node, and then restore the VMs.
 
 
Oh, thanks. I'll look at that for next time. I took the "backup then restore" route this time around. Simple and effective. Not quite the "technical" solution I was looking for like yours though. Thankfully I'm blessed by spare space and hypervisor capacity to fire up Proxmox Backup Server to make it easy on me.
 
The easiest option is to backup any VMs that matter from one of the nodes, then delete all of that node's VMs. Then you should be able to click in the UI on Datacenter and go to the storage section. Remove the storage for that node and create it again with a name that matches the other node, and then restore the VMs.
I ended up taking your advice. Fired up a VM on the extra capacity I had in a Hyper-V host, installed PBS on it, and the rest was just waiting for things to backup/restore. Deleted the poorly named storage on the newly empty node and went about my business. I did take the time while the node was empty to upgrade the PVE version from 8.4.1to 9.2.