NFS with constant stale file handles

[ruepel]

Hi

 

I seem to have a problem with nfs that I don't know how to troubleshoot. I have a media share on my Unraid system (6.6.3) that I am mounting via nfs on my plex server. That machine also does a few other things for me like importing music via beets. 

 

I have tried the following version

 

192.168.1.2:/mnt/user/media     /mnt/media      nfs     defaults      0       0
192.168.1.2:/mnt/user/media     /mnt/media      nfs     noatime,nodiratime      0       0
192.168.1.2:/mnt/user/media     /mnt/media		nfs 	defaults,intr,rw,vers=4,timeo=15,auto,async,nofail,bg		0		0

none of them actually resolve the issues. Plex is constantly rescanning the files and readding them to the library causing rebuilds of the metadata. Beets would oftentimes report stale file handles, and transcodes via ffmpeg would stop because of stale file handles. 

 

I have left the "tuneable" at default of 330 but experimented with a value of 0 to see if it would make difference but it didn't. 

 

I have also checked if the files on the server have been modified or accessed. None of the files that are reported to the plex server as "new" or "changed" have been created, changed, or accessed during my testing. That is files that have been in the library before and are reported to plex as new/changed.

 

My /etc/exports file reads:

"/mnt/user/media" -async,no_subtree_check,fsid=100 *(sec=sys,rw,insecure,anongid=100,anonuid=99,all_squash)

Do you have any idea what might cause those stale file handles and how I can get rid of them/ troubleshoot them more?

 

One workaround right is to mount the shares with CIFS which works fine but I would like to use NFS if I could.

 

thank you for your help!

[John_M]

The early Unraid 6.6 releases had NFS issues. Upgrade to 6.6.5 and test again.