[Mailman-Developers] Re: your mail

Scott scott@chronis.pobox.com
Thu, 28 May 1998 23:00:06 -0400


On Thu, May 28, 1998 at 10:29:17PM -0400, Corbett Klempay wrote:
| Does anyone have any quantitative (or maybe even non-quantitative) idea of
| Mailman's performance, especially compared to direct competitors such as
| Majordomo?  

not that i know of.

| Do we know of any very high volume lists being run on Mailman
| (or is this even a good idea?).  I 


The highest volume that i have heard of is ~250.  Over the next number
of weeks, i plan to give mailman a test run for some large lists.  If
i'm lucky i'll get some lists with thousands of subscribers.  

I think large lists with mailman is a good idea, and that mailman's
programming is structured in a way that if it cannot handle very large
lists as is, it will be easy to accomodate the needed changes.  I am
certain it will become quite scalable over the summer :)

| imagine that something like ezmlm would
| probably obliterate both Majordomo as well as Mailman as far as
| performance, but I don't see the average list needing the kind of fiendish
| performance ezmlm provides (don't me wrong; ezmlm has a lot to recommend
| it); most people could (and would) trade ezmlm's speed for Mailman's slick
| web interface.  But what I'm wondering is do we know how big this
| performance tradeoff is?

As an administrator of a very high volume list site, i can assure you
that the efficiency of the MTA is so much more important than the
efficiency of the mailling list manager that the efficiency of the
mailling list manager can *almost* be ignored. 

As i see it, mailman's efficiency is OK now, and can and will be made
more and more efficient by implementing better handling of locking the
lists, general code cleanup, and perhaps a client-server model.

Without any hard data, in many ways it seems like mailman will
outperform majordomo with most list operations in loaded circumstances
simply because it deals with a lot less file io, which is one the main
things that tends to slow down a host that is delivering lots of mail.

Also, with the upcoming use of deliveries via smtp, it will be much
easier to have the web interface for mailman on one machine and the
outgoing deliveries happen on another machine.  To my knowledge, most
really large listservs do this, and it would be quite difficult to get
to work with majordomo.  

I'm not familiar with ezmlm, but I know that no matter how efficient
it is, it almost certainly won't make a noticable difference compared
to differences produced by how an MTA handles mail, and if it's
difficult to have delivery occur on a separate machine with ezmlm,
then it can't scale as well as a mailling list manager that allows for
that separation.


| 
| Also, I saw that there was a mention on the TODO list about adding support
| for other MTAs (such as qmail).  Has anyone implemented this on their own
| yet?  Would providing support for qmail affect (enhance) performance much?

the upcoming smtp changes will make any MTA work for delivery.  i
don't think anyone has made something that will manage aliases for
qmail, though.

| I'm trying to (seeing how smooth Mailman is) see what kind of potential
| user base there will be; I'm wondering at what point the list becomes too
| speed-critical to allow use of Mailman.  Any comments?

Once mailman is past it's beta stage, i'm willing to bet that it will
handle large lists very well.  In the mean time, there may well be
some unforeseen glitches.

scott