At 10:37 AM -0500 1/29/01, Michael Lorrey wrote:
>Natasha's message here seems to have locked up download from the linux
>mail server to my nestcape mail reader, it seems truncated header
>information is the culprit.
>Received: from tick.javien.com ([188.8.131.52])
>com>; Sun, 28 Jan 2001 19:18:53 -0500
>ing: tick.javien.com: majordom set sender to
>firstname.lastname@example.org using email@example.com
>t: PARTY: Release of Primo 3M+
> * * * * *
> * * * * *
Mike, you have repeatedly complained about extropian messages locking
up your server. I don't think anybody else is having that problem.
I think the problem is in your server somewhere. If extropians is
the most verbose list to which you subscribe, it just may be
statistically more likely to be in the queue when the problem occurs.
The above message came through to me without truncated header
information. The specific headers you see truncated were not
truncated in my copy. Therefore, it seems to be coming out of the
ExI servers unbroken and is being garbled elsewhere on its path to
you. Note that the truncation does not seem to be positional or
line-based, but is similar to what a lack of flow-control would
produce. My first guess would be that the connection you use for
downloading mail from your linux server is not performing
flow-control correctly. This would probably corrupt the message and
hang your connection at the same time. If you are using a modem,
make sure the hardware flow-control is enabled. If you must use
software flow-control, try increasing your buffer size to something
Here are my suggestions for isolating the problem:
Copy the offending message and send it to yourself, does it lock up
again? If so, something in the message is the trigger and can be
repeatedly demonstrated. If not, the message itself is not causing
the crash but may have just happened to be in the queue at the time
of the crash.
If the message is the trigger, send it to other linux mail servers or
netscape mail readers to see if they likewise crash. If so, it is a
common problem. If not, the problem is isolated to your particular
configuration. This could be a settings problem, an environment
problem (such as disk space, memory size, number of processes), or
some random corrupted file.
Note that bad data should never crash a data handler such as e-mail.
If this were allowed, the server could be deliberately crashed by
hackers or would be unstable whenever bad data is received. In
general, such systems must be fixed even if the data is the trigger.
The complete text of the message is reproduced below. This will give
you a complete copy to play with in addition to your corrupted copy.
Received: from tick.javien.com ([184.108.40.206])
by spool2.valueweb.net (8.9.2/8.9.2) with ESMTP id TAA86480
for <extropians@HarveyNewstrom.com>; Sun, 28 Jan 2001 19:02:23 -0500
Received: (from majordom@localhost)
by tick.javien.com (8.9.3/8.9.3) id QAA09025
for extropians-outgoing; Sun, 28 Jan 2001 16:25:38 -0700
X-Authentication-Warning: tick.javien.com: majordom set sender to
firstname.lastname@example.org using -f
X-Mailer: Windows Eudora Pro Version 3.0 (32)
Date: Sun, 28 Jan 2001 15:23:30 -0800
From: Natasha Vita-More <email@example.com>
Subject: PARTY: Release of Primo 3M+
Content-Type: text/plain; charset="us-ascii"
* * * * *
* * * * *
-- Harvey Newstrom <http://HarveyNewstrom.com>
This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:56:26 MDT