From owner-freebsd-isp Wed Nov 19 05:58:32 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id FAA17205 for isp-outgoing; Wed, 19 Nov 1997 05:58:32 -0800 (PST) (envelope-from owner-freebsd-isp) Received: from nak.myhouse.com (nak.myhouse.com [209.70.45.162]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id FAA17187 for ; Wed, 19 Nov 1997 05:58:17 -0800 (PST) (envelope-from zoonie@nak.myhouse.com) Received: (qmail 2430 invoked by uid 1000); 19 Nov 1997 13:59:55 -0000 Date: Wed, 19 Nov 1997 08:59:55 -0500 (EST) From: zoonie Reply-To: zoonie To: Ulf Zimmermann cc: danny@panda.hilink.com.au, scot@poptart.org, akl@wup.de, amr@wup.de, isp@FreeBSD.ORG Subject: Re: RIP vs. OSPF In-Reply-To: <199711190640.WAA22585@Gatekeeper.Alameda.net> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-isp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Tue, 18 Nov 1997, Ulf Zimmermann wrote: > Aehm. OSPF has in a stable network less overhead (network traffic and CPU > wise) then RIP. OSPF only sends HELLOs, while Rip sends every 30 seconds > the whole routing table. If you have a flapping interface, then OSPF can > produce much traffic, because of the flooding. But depending how you design > your network, this is catched at Area borders. > you are correct ulf and i realized that i wasn't that clear on what i meant by overhead after i sent the message, it was getting late for me and i was really tired. i thought that somebody would ding me on that. i guess that it comes down to preference. if you don't have a large network should you really bother to create areas or have one area and run the risk of the flooding and computing the SPF algorithm everytime something changes or if an interface flaps. obviously if the network is stable this won't happen all the time and you do want stability. my preference would be for a link state routing protocol also but for some networks i just don't see the point. if you don't have a lot of traffic going through and the network is not very large and you don't have redundant paths what difference do a few packets every 30 seconds and some extra CPU cycles make? to me it's not a big deal, to others it may be. for the myhouse network i wanted to use VLSM to make better use of our current address space. i had the same choices. since the network isn't very large, i can live with the 30 second updates and i didn't see any good reason to run OSPF over RIPv2. so i chose RIPv2 and it still gets the job done which is the important thing to me.