From owner-cvs-all Fri Oct 13 7:57:26 2000 Delivered-To: cvs-all@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id 080CD37B66D; Fri, 13 Oct 2000 07:57:22 -0700 (PDT) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id KAA55959; Fri, 13 Oct 2000 10:57:10 -0400 (EDT) (envelope-from wollman) Date: Fri, 13 Oct 2000 10:57:10 -0400 (EDT) From: Garrett Wollman Message-Id: <200010131457.KAA55959@khavrinen.lcs.mit.edu> To: Don Lewis Cc: developers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/etc inetd.conf In-Reply-To: <200010130737.AAA05256@salsa.gv.tsc.tdk.com> References: <200010121857.e9CIvAi30686@earth.backplane.com> <20001013031655.K37870@jade.chc-chimes.com> <200010130737.AAA05256@salsa.gv.tsc.tdk.com> Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG < said: > In the past there have been widespread deployments of network software > with example cron entries. Most users would just configure it to run > at the default time, and the poor server at the other end would melt > down. Just ask anyone who's ever run a POP server for an office full of Eudora users. `xload' looks like a veritable sawtooth with those spikes exactly five minutes apart. Work done by Lixia Zhang and others has demonstrated that this can and will happen, in sufficiently large networks, *even in the absence of synchronized clocks*, because events like this are self-synchronizing. The only solution is to randomize the interval. -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message