Shared Storage (NFS/PBS) go offline at random, but stay online for other nodes

[mcn19]

Hello, I'm new to Proxmox and I've setup both an NFS and PBS for ISOs and Backups respectively. I've been encountering an issue where seemingly at random, both "backup" and "nfs-iso" enter into a "status: unknown" state on one or two of my nodes and do not reconnect. Sometimes they reconnect with a reboot, or by completely removing and readding them from Datacenter > Storage.

I can ping both the IP of the NFS and PBS from the node, so they're talking, but just won't mount or show as 'status:available' in the UI (PVE 8.3.0).

The following is my /etc/pve/storage.cfg file:

        path /var/lib/vz
        content backup,iso,vztmpl

lvmthin: local-lvm
        thinpool data
        vgname pve
        content images,rootdir

nfs: nfs-iso
        export /mnt/isos
        path /mnt/pve/nfs-iso
        server 192.168.4.106
        content iso
        prune-backups keep-all=1

pbs: backup
        datastore pbs-backup
        server 192.168.4.105
        content backup
        fingerprint
        username root@pam

journalctl -xe just tells me that it's not online and it can't connect, despite being able to ping out:

Feb 13 12:06:24 MiniServer2 pvestatd[984]: storage 'nfs-iso' is not online
Feb 13 12:06:31 MiniServer2 pvestatd[984]: backup: error fetching datastores - 500 Can't connect to 192.168.4.105:8007

Any ideas on what to do here? Proxmox is all new to me so I'm pretty lost.

 

[dj423]

Have you tried to connect to the specific port by chance, with something like netcat? ie: nc -zv 192.168.4.105 8007
and I thought nfs listens on 2049, but I could be wrong.

Can also try: pvesm scan nfs 192.168.4.105 - just to see what is showing available. and mount | grep 'type nfs'

 

[mcn19]

Have you tried to connect to the specific port by chance, with something like netcat? ie: nc -zv 192.168.4.105 8007
and I thought nfs listens on 2049, but I could be wrong.

Can also try: pvesm scan nfs 192.168.4.105 - just to see what is showing available. and mount | grep 'type nfs'

Thanks for the reply, I did try to connect to the ports initially but I believe I hit the same "can't connect" issues.

I opted for the nuclear option last night and migrated my vms then reinstalled proxmox on each node, so far the issue hasn't come back, but will troubleshoot further if it does