Return to BSD News archive
Newsgroups: comp.unix.bsd.freebsd.misc
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!news.mira.net.au!news.vbc.net!garlic.com!news.scruz.net!kithrup.com!news.Stanford.EDU!bloom-beacon.mit.edu!news.mathworks.com!uunet!in2.uu.net!zorch!zorch.sf-bay.org!scott
From: scott@zorch.sf-bay.org (Scott Hazen Mueller)
Subject: Re: Minimum machine for nameservice?
Reply-To: scott@wvs.com
Sender: usenet@zorch.SF-Bay.ORG (Charlie Root)
Organization: Worldview Systems
Message-ID: <Dzr4BH.7tt@zorch.SF-Bay.ORG>
References: <chad-1710962353430001@sverige.pengar.com> <548n0e$61f@shimon.netmedia.net.il> <54dg0h$3gs@uriah.heep.sax.de> <Dzn2rt.2y4@zorch.sf-bay.org> <54jkmf$k3o@uriah.heep.sax.de>
X-Nntp-Posting-Host: localhost.sf-bay.org
Date: Wed, 23 Oct 1996 23:09:16 GMT
Lines: 14
>Even then, it's certainly that you're referring to busy name servers,
>while i am writing about mostly idle servers (and the originator did
>clarify that it won't be a very busy one at all).
Only moderately busy, and it probably doesn't matter. My experiences suggest
that all it takes is one good burst that goes unanswered to start a snowball
effect that causes swap thrashing. For instance, if you run a bogus mail
gateway with bad timeout values. If a mail message comes in and the
nameserver takes too long to get pulled off of swap, the message winds up
re-queued and the same thing happens next time. Pretty soon, you log in and
find thousands of messages in your queue and dozens of unhappy sendmails.
\scott