AF:
NF:0
PS:10
SRH:1
SFN:
DSR:
MID:<20080918173717.67fe0e63@ripper.onstor.net>
CFG:
PT:0
S:andy.sharp@onstor.com
RQ:
SSV:onstor-exch02.onstor.net
NSV:
SSH:
R:<john.rogers@onstor.com>,<larry.scheer@onstor.com>,<sandrine.boulanger@onstor.com>,<brian.stark@onstor.com>
MAID:1
X-Sylpheed-Privacy-System:
X-Sylpheed-Sign:0
SCF:#mh/Mailbox/sent
X-Sylpheed-End-Special-Headers: 1
Date: Thu, 18 Sep 2008 17:38:36 -0700
From: Andrew Sharp <andy.sharp@onstor.com>
To: John Rogers <john.rogers@onstor.com>
Cc: Larry Scheer <larry.scheer@onstor.com>, Sandrine Boulanger
 <sandrine.boulanger@onstor.com>, Brian Stark <brian.stark@onstor.com>
Subject: plan for debugging exim hangs on dogfood
Message-ID: <20080918173836.73527af4@ripper.onstor.net>
Organization: Onstor
X-Mailer: Sylpheed-Claws 2.6.0 (GTK+ 2.8.20; x86_64-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

Step 1. Would be great if there was a dev machine that could be used
to try and reproduce this.  Facts seem to indicate that it needs to be
a cluster.

Step 1a.  If we can find a QA machine that demonstrates the same
symptoms, we can use that instead of dogfood.

Step 2. I would like a system compare to be done on dogfood against a
build of the correct version.  I want Larry to do the build and the
system compare (or at least be present) because I know he will get
everything precisely nailed down like the right source checked out, the
system compare command with the right arguments and so forth, and can
interpret the results.

Step 3. I would like to install a special version of exim on dogfood
(or equivalent QA machine if there is one).  First I would like to just
see if the problem reproduces with this special version I have put
together.  If so, then I will want to run exim with gdb to debug some
of the code paths.  I would need to be on the machine for about a day,
but it shouldn't effect the operation of the machine except for
autosupport emails.

I don't anticipate it being necessary, but I might need to NFS mount
some source for gdb.

Exit strategy: restore the original version of exim unless the special
version does not demonstrate the problem.  Unmount any extraneous NFS
mounts.

