I'm facing a strange issue with my NFSv4 setup. I've got multiple servers trying to mount NFS shares from a single exporting server, but only one server is able to do it without issues. The other servers are generating errors like 'Connection refused' and 'Device or resource busy' for both IPv4 and IPv6 addresses. I've checked the /etc/fstab files on all servers, and they all have the same mount options. I'm not seeing anything in the logs on the exporting server that indicates a problem. Does anyone have an idea of what's going wrong or how I could troubleshoot this issue?
3 Answers
If you’re getting 'Device or resource busy', it could point to something else using that path while you're trying to mount. Try ensuring no other process is holding the filehandle during the mount attempt.
When you see 'Connection refused' errors, it usually means the NFS daemon isn't listening on the expected interfaces. Check your firewall settings and ensure that your NFS daemon is correctly bound to the IPv6 interface. If your NFS server is starting before the network is fully up, that could also be a factor.
Interesting thought! My firewall hasn't changed, and it worked fine on other servers. I might have to investigate if the NFS daemon is starting correctly in relation to the network setup.
It's odd that only one server is able to mount the shares properly. Usually, I've found that this problem occurs when the NFS clients aren't added correctly to the /etc/exports file or if you've duplicated the fsid. But since it fixed itself, it could have been a temporary glitch—maybe a networking thing?
That's a valid point. This isn't a new export, so I'm not sure why it would suddenly cause issues if nothing has changed.
Yeah, that makes sense. I haven't noticed anything, but I'll keep an eye out for any processes that might interfere next time.