Need Help to Fix Stale NFS File Handle

[Matt Graham]

From: Mark Phillips <mark at phillipsmarketing.biz>
> On Fri, Feb 25, 2011 at 8:28 AM, Jason Holtzapple <ml at bitflip.net> wrote:
>> On 02/25/2011 08:23 AM, Mark Phillips wrote:
>>> channel for the photo disk died, but the os kept on running.
>>> I replaced the controller card, and now both drives are running.
>>> However, I am getting some stale NFS file handles on some of
>>> the images on the photo drive. I don't use NFS on this machine

>>> hammerhead:/home/mark# du -hs /backups
>>> du: cannot access `/backups/gallery3/var/resizes/11-01-2010/2010:11:01
>>> 08:11:50 295.JPG': Stale NFS file handle
>> This is strange - you shouldn't be getting these errors unless you are
>> an NFS client. Double check your /etc/fstab and /etc/exports files.

Yeah, what Jason said.  "Stale NFS file handle" is only supposed to happen if
you have an NFS filesystem mounted, and that mount gets interrupted by the NFS
server barfing or the network dying.  What's the output from "mount" and "ps
auxw | grep nfs"?

> hammerhead:/home/mark# grep nfs /etc/fstab
> hammerhead:/home/mark#
> hammerhead:/home/mark# egrep -v '^#' /etc/exports
> /home/mark/vmware    192.168.25.0/255.255.255.0(rw)

So this machine was acting as an NFS server at some point?  Hm.  nfsd running
on the machine shouldn't cause this.  Is there anything that looks weird in
the output from "dmesg | grep -i nfs"?  

There are historical reasons why "NFS" expands to "Nightmare File System"...
the thing is, it performs faster than Samba or Coda, so it gets used, even
though it has a number of annoying problems.  Most of the kernel panics I've
seen have been caused by flaky hardware--but NFS and "don't install this
kernel module, it's not even beta yet" are tied for second.

-- 
Matt G / Dances With Crows
The Crow202 Blog:  http://crow202.org/wordpress/
There is no Darkness in Eternity/But only Light too dim for us to see