could not activate storage 'local-zfs', zfs error: cannot import 'rpool': no such pool available (500) after cluster join

[johniem20]

Hi, I am new to this so excuse my ignorance. I had two Proxmox nodes unjoined running seperate VMs. I decided to joining them together in a cluster. I found out you can't join nodes that have a pre-existing VM to a cluster so I:
1. stopped the VM,
2. backed it up,
3.move the VMs descriptors within /etc/pve/nodes/[NODENAME]/qemu-server to /home/_cluster_bck
4. join the cluster
5. copy the backup VMs descriptors back to /etc/pve/nodes/[NODENAME]/qemu-server

Upon joining it has been complaining about ZFS pool with the error could not activate storage 'local-zfs', zfs error: cannot import 'rpool': no such pool available (500). From looking it up I think I had this as LVM(based on backup config below) but could not be certain. Source for google on that: (https://www.reddit.com/r/Proxmox/co...fs_rpool_not_available_after_joining_cluster/)

The config for the backup is as follows:
#<div align='center'><a href='https://helper-scripts.com' target='_blank' rel='noopener noreferrer'><img src='https://raw.githubusercontent.com/tteck/Proxmox/main/misc/images/logo-81x112.png'/></a>
#
# # Home Assistant OS
#
# <a href='https://ko-fi.com/D1D7EP4GF'><img src='https://img.shields.io/badge/&#x2615;-Buy me a coffee-blue' /></a>
# </div>
agent: 1
bios: ovmf
boot: order=scsi0
cores: 2
cpu: host
efidisk0: local-lvm:vm-100-disk-0,efitype=4m,size=4M
localtime: 1
memory: 4096
meta: creation-qemu=8.1.5,ctime=1714494921
name: haos12.2
net0: virtio=02:2F:F6:F6:9B:CF,bridge=vmbr0
onboot: 1
ostype: l26
scsi0: local-lvm:vm-100-disk-1,cache=writethrough,discard=on,size=32G,ssd=1
scsihw: virtio-scsi-pci
smbios1: uuid=ebfcda7a-b1be-44a1-a6d5-adf4a0063252
tablet: 0
tags: proxmox-helper-scripts
usb0: host=1a86:55d4
vmgenid: 777f0445-d074-41cc-84c1-8af3dc7457e0

#qmdump#map:efidisk0:drive-efidisk0:local-lvm:raw:
#qmdump#map:scsi0:drive-scsi0:local-lvm:raw:
____________________________________________________________________________________


Is there a way I can rectify this without loosing my data? Can I restore my backed up up VM to any of my nodes?

 

[Chris]

Upon joining it has been complaining about ZFS pool with the error could not activate storage 'local-zfs', zfs error: cannot import 'rpool': no such pool available (500).

Given this error message and that you state that you have LVM on that node, my guess is that this particular node has no zpool set up. You can go to Datacenter > Storage > local-zfs > Edit and limit the storage to the nodes it is available on. Same for the local-lvm which most likely does not exist on the nodes setup using ZFS.

I had two Proxmox nodes unjoined running seperate VMs. I decided to joining them together in a cluster.

You should definitely have a look at how to setup an external voter for the cluster, as otherwise you will run into quorum issues if one of the nodes acts up or is shut down, see https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_corosync_external_vote_support