Great Circle Associates Majordomo-Workers
(September 1996)
 

Indexed By Date: [Previous] [Next] Indexed By Thread: [Previous] [Next]

Subject: Re: Majordomo Bombing
From: Jason L Tibbitts III <tibbs @ hpc . uh . edu>
Date: 25 Sep 1996 13:36:34 -0500
To: Evan Champion <evanc @ synapse . net>
Cc: majordomo-workers @ GreatCircle . COM
In-reply-to: Evan Champion's message of Wed, 25 Sep 1996 00:27:36 -0400
References: <3248B4B8.F8@synapse.net>

>>>>> "EC" == Evan Champion <evanc@synapse.net> writes:

EC> I just had a rather unfortunate incident where one of my list admins
EC> sent through something like 300 individual subscribe/unsubscribe
EC> messages as well as 4 or 5 postings to a 20000 user mailing list, all
EC> at once.

Read my message last night about torturing majordomo.

EC> The basic problem, as I see it, is that Majordomo has no concept of
EC> system resources.  I mean, it assumes everyone has enough CPU, memory
EC> and I/O capacity to handle any job, no matter how large.

Why should it?  It sits between two sendmail invocations.  Sendmail won't
even deliver the mail (i.e. invoke wrapper in the first place) if the load
is too high.  It won't even accept new connections if the load is really
high.  If your system cannot handle the loads that it's configured at, then
reduce them.

EC> So, the question is then, is there anything that can be done to
EC> minimise the stress Majordomo causes on a system.  I have been writing
EC> this message for the last 2 hours and I still haven't come up with any
EC> ideas I really like.

We've been over this several times.  I've stated the direction I think 2.x
should take, which involves constantly running daemons that keep the
necessary state to minimize startup time.
-- 
      Jason L. Tibbitts III - tibbs@uh.edu - 713/743-8684 - 221SR1
System Manager:  University of Houston High Performance Computing Center
                1994 PC800 "Kuroneko"      DoD# 1723



References:
Indexed By Date Previous: Majordomo 1.94 Beta 4. tiny knob tweaking.
From: Chan Wilson <cwilson@slurp.neu.sgi.com>
Next: Patch for 1.94 to avoid starvation when looping for lock
From: John Gilmore <gnu@toad.com>
Indexed By Thread Previous: Re: Majordomo Bombing
From: Brent Chapman <Brent@GreatCircle.COM>
Next: Preliminary patch to call beta 4.
From: Chan Wilson <cwilson@slurp.neu.sgi.com>

Google
 
Search Internet Search www.greatcircle.com