[299] in arla-drinkers
Re: Machine hanging with arla-0.12
daemon@ATHENA.MIT.EDU (Dr A V Le Blanc)
Fri Oct 2 05:38:01 1998
From owner-arla-drinkers@stacken.kth.se Fri Oct 02 09:38:00 1998
Return-Path: <owner-arla-drinkers@stacken.kth.se>
Delivered-To: arla-drinkers-mtg@bloom-picayune.mit.edu
Received: (qmail 29375 invoked from network); 2 Oct 1998 09:37:59 -0000
Received: from unknown (HELO sundance.stacken.kth.se) (130.237.234.41)
by bloom-picayune.mit.edu with SMTP; 2 Oct 1998 09:37:59 -0000
Received: (from majordom@localhost)
by sundance.stacken.kth.se (8.8.8/8.8.8) id LAA27362
for arla-drinkers-list; Fri, 2 Oct 1998 11:31:52 +0200 (MET DST)
Received: from serenity.mcc.ac.uk (serenity.mcc.ac.uk [130.88.200.93])
by sundance.stacken.kth.se (8.8.8/8.8.8) with ESMTP id LAA27358
for <arla-drinkers@stacken.kth.se>; Fri, 2 Oct 1998 11:31:48 +0200 (MET DST)
Received: from cguhpc.cgu.mcc.ac.uk ([130.88.201.14] ident=zlsiial)
by serenity.mcc.ac.uk with esmtp (Exim 1.92 #2)
id 0zP1Yx-0006Nf-00; Fri, 2 Oct 1998 10:31:47 +0100
Received: (from zlsiial@localhost)
by cguhpc.cgu.mcc.ac.uk (8.7.6/8.8.4)
id KAA10514; Fri, 2 Oct 1998 10:31:46 +0100 (BST)
Message-ID: <19981002103146.B10490@afs.mcc.ac.uk>
Date: Fri, 2 Oct 1998 10:31:46 +0100
From: Dr A V Le Blanc <LeBlanc@mcc.ac.uk>
To: Assar Westerlund <assar@sics.se>
Cc: arla-drinkers@stacken.kth.se
Subject: Re: Machine hanging with arla-0.12
Reply-To: Dr A V Le Blanc <LeBlanc@mcc.ac.uk>
References: <19981001162019.A10086@afs.mcc.ac.uk> <5lk92kwako.fsf@assaris.sics.se> <19981001173850.A10121@afs.mcc.ac.uk> <5l90j0w6oy.fsf@assaris.sics.se> <5llnmzhavg.fsf@assaris.pdc.kth.se>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-Mailer: Mutt 0.93.1i
In-Reply-To: <5llnmzhavg.fsf@assaris.pdc.kth.se>; from Assar Westerlund on Fri, Oct 02, 1998 at 05:59:47AM +0200
Sender: owner-arla-drinkers@stacken.kth.se
Precedence: bulk
On Fri, Oct 02, 1998 at 05:59:47AM +0200, Assar Westerlund wrote:
> So I've been running 2.1.123 + the debugging malloc patches that I
> sent you earlier (with the obvious typo fixed :-) and the appended
> patch which add a few xfs_free's and it does seem to be working.
...
> Anyways, please try this patch (I included the
> previous one as well) and tell me if your hangs still occur.
I'm sorry; I seem to have missed the second patch, but since you
included it in this third one (along with the first patch), I've
now put it in. Then I copied the 800-odd files from /arla to a
local disk (580mb).
Symptoms: performance is incredibly better, there are no hangs, and a
job which took several hours with /afs mounted using NFS now completes
in about 10 minutes, leaving the system normal afterwards. There are
two unfortunate side effects: many of the files copied by this job are
now coming across corrupt. Moreover, a long series of cmps to find out
how many files got corrupted causes the system to panic. (These are
"for i in `cat files`;do cmp $localdir/$i /arla/xxx/$i;done" jobs, so
they are reading all the files from arla a second time.)
Clearly progress is being made. Thanks again, Assar, for the
very quick response.
-- Owen
LeBlanc@mcc.ac.uk
