Storage Question after Joining a Node to Cluster

[bcops]

Hello,

I just created a cluster comprising of the following servers:

1. Cluster Name: cluster-a
   1. server-a
      1. Proxmox 6.2-4
      2. ZFS:

         1. root@server-a:/storagez3/storage# zfs list
            NAME USED AVAIL REFER MOUNTPOINT
            rpool 72.9G 142G 104K /rpool
            rpool/ROOT 72.9G 142G 96K /rpool/ROOT
            rpool/ROOT/pve-1 72.9G 142G 72.9G /
            rpool/data 96K 142G 96K /rpool/data
            storagez3 20.7T 2.36T 363G /storagez3
            storagez3/backup 279K 2.36T 279K /storagez3/backup
            storagez3/images 279K 2.36T 279K /storagez3/images
            storagez3/storage                 20.3T  2.36T      279K  /storagez3/storage

   2. server-b
      1. Proxmox 6.2-4
      2. ZFS

         1. root@server-b:~# zfs list
            NAME USED AVAIL REFER MOUNTPOINT
            rpool 929M 214G 104K /rpool
            rpool/ROOT 928M 214G 96K /rpool/ROOT
            rpool/ROOT/pve-1 928M 214G 928M /
            rpool/data          96K   214G       96K  /rpool/data

After succesfully joining the cluster, server-b generates the following error output:

root@server-b:~# journalctl -f
May 20 18:22:00 server-b systemd[1]: Starting Proxmox VE replication runner...
May 20 18:22:00 server-b systemd[1]: pvesr.service: Succeeded.
May 20 18:22:00 server-b systemd[1]: Started Proxmox VE replication runner.
May 20 18:22:08 server-b pvestatd[1568]: zfs error: cannot open 'storagez3': no such pool
May 20 18:22:08 server-b pvestatd[1568]: zfs error: cannot open 'storagez3': no such pool
May 20 18:22:08 server-b pvestatd[1568]: could not activate storage 'zfsimages', zfs error: cannot open 'storagez3': no such pool
May 20 18:22:08 server-b pvestatd[1568]: zfs error: cannot open 'storagez3': no such pool
May 20 18:22:08 server-b pvestatd[1568]: zfs error: cannot open 'storagez3': no such pool
May 20 18:22:08 server-b pvestatd[1568]: could not activate storage 'zfsstorage', zfs error: cannot open 'storagez3': no such pool

I originally attempted to join server-b to the cluster while it had a raidz2 named storagez2 which generated the exact same errors found in journalctl. I figured it was a problem caused by setting up a pool prior to joining the cluster. So while following the official documentation, I removed it from the cluster and re-installed Proxmox on server-b to join the cluster with a brand new setup.

Is there something missing that I do not see in the https://pve.proxmox.com/wiki/Cluster_Manager documentation? Is there some sort of storage sync that I need to turn off that I've missed from the documentation or forums? Any help would be greatly appreciated.

 

[wolfgang]

Hi,

the problem is in your case that you have not restricted storagez3 to server-a.
So if you join the cluster with server-b it is searching for zfs pool storagez3.
This storage is not available on the server-a.

You have 2 options.
1.) restrict storagez3 to server-a
2.) create a zfs pool with the same name on server-b.