From owner-freebsd-hackers Mon Aug 12 20:20:10 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id UAA10864 for hackers-outgoing; Mon, 12 Aug 1996 20:20:10 -0700 (PDT) Received: from parkplace.cet.co.jp (parkplace.cet.co.jp [202.32.64.1]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id UAA10858 for ; Mon, 12 Aug 1996 20:20:07 -0700 (PDT) Received: from localhost (michaelh@localhost) by parkplace.cet.co.jp (8.7.5/CET-v2.1) with SMTP id DAA15061; Tue, 13 Aug 1996 03:19:55 GMT Date: Tue, 13 Aug 1996 12:19:55 +0900 (JST) From: Michael Hancock To: Brandon Gillespie cc: freebsd-hackers@FreeBSD.ORG Subject: Re: Load-balancing box In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Mon, 12 Aug 1996, Brandon Gillespie wrote: > On Tue, 13 Aug 1996, Michael Hancock wrote: > > > > webserverA is what DNS advertises. webservers 1 to 3 have mirrored > > content. > > > > When the redirector box starts up all servers are given zero cost requests > > are redirected on a least cost basis with round-robin on identical costs. > > (This is just an example of a distribution policy) > > > > Servers that don't respond are assigned infinite cost and a > > back-in-service algorithm can be used to get the rebooted server back in > > the pool. > > Wouldn't it be simpler to hack the name daemon to do the load balancing, > so when they lookup 'www.xxx.yyy' it picks a machine and directs them to > the IP for 'wwwX.xxx.yyy'? From that point on you dont care what they are > doing. I know VMS can cluster like this, without a problem (through > MultiNet). Are you talking about round robin DNS? This doesn't work very well. You don't want the IP to be cached. It difficult to load balance thru DNS. How do you deal with servers that go down? Mike