[SOLVED] Proxmox 8 can't attache NFSv4 share (Synology) to both nodes, only one can connect.

[psybear]

Hi everyone,

I have an NFSv4 share setup on Synology I'm trying to add to 2 nodes Proxmox v8 (pve1 & pve2).

I've used the Proxmox Cluster admin UI to add the share.

pve1 connects no problem, however pve2 times out.

Detailed setup info can be reviewed here -> https://gist.github.com/sergeifilippov/674d78377e26001ad9183665a4e5e7e2

 

[bbgeek17]

1) you have duplicate IP in your table
2) you nfsscan two different IPs

You should resolve any conflicts, or provide correct information. If you are getting timeouts, try basic network troubleshooting: ping, arp, nc, etc

---

Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

 

[psybear]

1) you have duplicate IP in your table
2) you nfsscan two different IPs

You should resolve any conflicts, or provide correct information. If you are getting timeouts, try basic network troubleshooting: ping, arp, nc, etc

---

Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

Sorry that was a typo

 

[bbgeek17]

Sorry that was a typo

Ok, great, those happen. As you can see from the output the "pvesm" command wraps around common Linux NFS utilities, ie "showmount" and "rpcinfo".

You need to figure out why that general Linux command is timing out. Things you can look at and optionally provide to the community:
a) "ip a" from each node
b) icmp ping from each node to your nfs server
c) direct "showmount" and "rpcinfo" from each node
d) you may want to use "nc host port" to test connectivity
e) if all else fails, tcpdump during "pvesm status" or "nfs scan".

Everything so far points to a networking issue, rather than hypervisor problem.

PS "dmesg" is not helpful when troubleshooting Userland issues, you may want to "journalctl -n 100" or "journalctl -f" in a second terminal session.

PPS keep in mind that NFS UDP has been disabled in Linux Kernel at some point, you may want to make sure that you are using TCP

PPPS output of "cat /etc/pve/storage.cfg" is also helpful , even though basic NFS tools are not working.

---

Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

 

[psybear]

Thanks @bbgeek17

Super helpful things to try! Will go away and give it a go and report back.

 

[psybear]

Found the issue in the network. Not a proxmox issue in end.

Thanks for all the comments