Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!spool.mu.edu!uwm.edu!www.nntp.primenet.com!nntp.primenet.com!enews.sgi.com!news.sgi.com!news1.best.com!nntp1.best.com!usenet
From: dillon@flea.best.net (Matt Dillon)
Newsgroups: comp.mail.sendmail,comp.mail.smail,comp.unix.bsd.freebsd.misc
Subject: Re: Sendmail vs. Smail...
Date: 16 Dec 1996 20:07:04 GMT
Organization: BEST Internet Communications, Inc.
Lines: 41
Message-ID: <594a58$g9a@nntp1.best.com>
References: <57tf61$gq7@raven.eva.net> <58rvbf$r6t@news.fsu.edu> <1996Dec1510.41.00.4656@koobera.math.uic.edu> <1996Dec16.152941.1706@isac.hces.com>
NNTP-Posting-Host: flea.best.net
Xref: euryale.cc.adfa.oz.au comp.mail.sendmail:35253 comp.mail.smail:2725 comp.unix.bsd.freebsd.misc:32691
:In article <1996Dec16.152941.1706@isac.hces.com>,
:Simon Casady <cap@hces.com> wrote:
:>D. J. Bernstein (djb@koobera.math.uic.edu) wrote:
:>: The problem with your attitude is that you end up trapped by your own
:>: ignorance. You start out not understanding something---e.g., why an
:>: extra RCPT can cost so much time---and you refuse to check your mental
:>: model against reality. So you never learn.
:>
:>So tell us, please, why does the RCPT cost so much time? Don't just
:>repeat that it does but show us why.
:>
:>--
:>Simon Casady Pace Health Management Systems
:>cap@hces.com voice (515)-222-1717
:>casady@acm.org fax (515)-222-1716
He's talking about the DNS delay... sendmail wants to run DNS
lookups for each RCPT. If you are mailing a destination with
broken name servers or a broken network, this will create a delay.
The delay is nominally on the order of 60 seconds per broken
address (for each *different* broken address.. the same broken
address several times only results in a single DNS timeout's delay).
However, adjusting DNS timeouts in sendmail is trivial. I just leave
ours alone, it works fine with the 100,000+ messages/day we run...
but if you really want to, you can reduce this from 2 30 second
attempts to 1 15 second attempt (i.e. four times faster). If you
are running named locally, this will work just fine because named
will cache any valid responses even if they come in after sendmail
gives up on it.
Running queue messages in parallel isn't really the right fix...
while it ultimately uses the same resources as running them in
serial, it only takes a few badly broken messages, spams, or
mail loops before you start wanting to turn off the 'feature'.
In a lightly loaded mail system, you can get away with it. In
a heavily loaded mail system it can potentially create glitches
big enough for your user base to notice.
-Matt