From jan@swi.swi.psy.uva.nl Fri Dec  7 09:50:47 2001
Received: from gollem.swi.psy.uva.nl (gollem.swi.psy.uva.nl [145.18.152.30])
	by swi.psy.uva.nl (8.10.2+Sun/8.11.2) with ESMTP id fB78olk06357
	for <prolog@swi.swi.psy.uva.nl>; Fri, 7 Dec 2001 09:50:47 +0100 (MET)
Received: (from jan@localhost)
	by gollem.swi.psy.uva.nl (8.11.3/8.11.3/SuSE Linux 8.11.1-0.5) id fB78olx15952
	for prolog@swi.swi.psy.uva.nl; Fri, 7 Dec 2001 09:50:47 +0100
Date: Fri, 7 Dec 2001 09:50:47 +0100
Message-Id: <200112070850.fB78olx15952@gollem.swi.psy.uva.nl>
From: Jan Wielemaker <jan@swi.psy.uva.nl>
Subject: Re: [SWIPL] Stale NFS file handle
To: prolog@swi.swi.psy.uva.nl
Phone: +31 - 20 - 525 6121

> I have encountered I/O error during file I/O:
> 
> ERROR: I/O error in read on stream $stream(322678) (Stale NFS file handle)
> 
> get_code/2 generated the above message. The file was NFS-mounted. I
> confirmed that the the file was mounted properly.
> 
> Could anybody advise me to avoid the problem?

As far as I recall, "Stale NFS file handle" means the file has gone on the
server.  It's 99.99% sure nothing to do with Prolog, but just with your setup.
One hint: if you monitor a file over NFS (not a good plan anyway) but the server
side updates the file by deleting it and writing a new one you may get this
trouble as the underlying inode changes.

It you still distrust Prolog itself, use a system-call tracer.  Called strace
on Linux, truss on Solaris, and maybe different on what you are using.

	--- Jan

