🚀 go-pugleaf

We've got a 4D/360 running Irix 4.0.5 as our main NFS server
(among other things).

I've been getting the error messages:

Jan 26 17:32:51 Uranium rpc.lockd[167]: get_fd: unable to cvt fh: Stale NFS file handle
Jan 26 17:33:51 Uranium last message repeated 29 times

etc., repeated roughly every two minutes for the past five
days (in /usr/adm/syslog).

I interpret this to mean that somebody's trying to open/read/write
a file which NFS thinks doesn't exist anymore, and is doing so
very persistently.

I've checked every workstation that NFS-mounts filesystems from
Uranium, done "umount -a" and then "mount -a" to try to flush things;
likewise I've killed (well, in a friendly fashion...) every user
process on every workstation which might have open files on the
server; tried "exportfs -ua" / "exportfs -a" on the server; all to no
avail.  I've rebooted the GatorBox which handles AppleShare-NFS
translation for us, also with no effect.

I've played with netstat and nfsstat, both on the server and on those
clients which support them, but haven't discovered anything useful.
I just got a copy of EtherPeek for my Mac - of course with no
documentation - but I don't see anything obvious in the traffic to/from
the server (of course if I knew what to look for it might be more
obvious...).

I'm about ready to reboot everything, but I'd like to know what's
causing this first.

So, can anyone tell me how to determine which machine/process is
causing this?  I'm setting up EtherPeek to capture traffic from
3 - 4 AM; what should i be looking for?

-- 
      Kurt Hillig
   Dept. of Chemistry      I always tell the     khillig@umich.edu
 University of Michigan     absolute truth    Telephone (313)747-2867
Ann Arbor, MI  48109-1055    as I see it.    hillig@chem.lsa.umich.edu