After an unscheduled reboot of our NFS server last night, a few of the clients reported "Stale NFS file handle".

All but one of these were fine after a reboot though one client is still giving errors after several reboots and manual umount/mounting of the filesystem.

What might be causing this be, and what could be done about it?

The OS is Ubuntu 10.04, with NFSv3.

link|improve this question
Is the error you are still seeing after reboots and remounts "stale file handle", or is it another error? – sciurus Mar 31 '11 at 13:41
feedback

1 Answer

up vote 0 down vote accepted

NFS locks can survive a reboot. They should eventually timeout, but sometime you need to shutdown NFS and clear the lock directory manually.

I believe I usually run into this issue when the export is not available when the system comes back up. If I remember right the locks are maintained in /var/lib/nfs/sm.

link|improve this answer
They did eventually clear. I can't tell if your suggestions of /var/lib/nfs/sm is right, but thanks. – jpnp Apr 21 '11 at 14:18
@jpnp: I believe the timeouts are quite long. I was never able to wait for them to timeout. – BillThor Apr 21 '11 at 15:38
feedback

Your Answer

 
or
required, but never shown

Not the answer you're looking for? Browse other questions tagged or ask your own question.