Re: Comments on RIP-II protocol and MIB specification.
Jeffrey C Honig <jch@nr-tech.cit.cornell.edu> Thu, 03 June 1993 18:29 UTC
Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa08005;
3 Jun 93 14:29 EDT
Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa08001;
3 Jun 93 14:29 EDT
Received: from atlas.xylogics.com by CNRI.Reston.VA.US id aa18122;
3 Jun 93 14:29 EDT
Received: by atlas.xylogics.com id AA24815 (5.65c/UK-2.1-930202);
Thu, 3 Jun 1993 14:28:14 -0400
Received: from MITCHELL.CIT.CORNELL.EDU by atlas.xylogics.com with SMTP
id AA25080 (5.65c/UK-2.1-930202); Thu, 3 Jun 1993 14:28:04 -0400
Received: from MITCHELL.CIT.CORNELL.EDU by mitchell.cit.cornell.edu
(4.1/1.34/Honig-1.3) id AA24672; Thu, 3 Jun 93 14:27:18 EDT
Message-Id: <9306031827.AA24672@mitchell.cit.cornell.edu>
To: Fred Baker <fbaker@acc.com>
Cc: ietf-rip@xylogics.com
Subject: Re: Comments on RIP-II protocol and MIB specification.
In-Reply-To: Message from fbaker@acc.com (Fred Baker) on
Thu, 03 Jun 1993 09:37:06 -0700.<9306031637.AA00889@saffron.acc.com>
Organization: Information Technologies/Network Resources;
Cornell University, Ithaca, NY
X-Mailier: MH-E [version 3.7+] MH [version 6.8]
Date: Thu, 03 Jun 1993 14:27:17 -0400
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: Jeffrey C Honig <jch@nr-tech.cit.cornell.edu>
You are assuming that Route Tags and Routing Domains are related, they are not. You basically have the concept of RouteTag correct. I can import external routes into a RIP domain from several different external attachement points, each using a different tag. At each external attachement point I can decide which, if any, external routes to export based on the tag. If I do not wish to announce any external routes, I can just announce those that have a tag of zero. Or maybe you want to tag each departments routes differntly and only announce some of those externally. Routing Domains solve a different problem. There are cases where two or more Administrative Domains (ADs) come together on one wire. Say we have a DMZ where there are two routers each from regional providers A and B (A1, A2, B1, B2) and a router from the national provider C (C1). The regionals use this network to connect the two halfs of their regional together as well for connecting to the other regional and to the national provider. So they want to be able to run an IGP between their routers (A1 to A2 and B1 to B2). If they use different IGPs everything is OK. If they want to use the same IGP, you need Routing Domains. To complicate things even further, say the national provider has agreed to listen to the IGPs of each of the regionals and implement their policy for them. So router C1 must be capable of listening to both IGPs and distinquishing between them. Gated has always implemented the capablility to unicast RIP updates to certain routers across a broadcast network. This is used when it is desired to restrict knowledge of certain routes to a subset of the routers on a given network. Say the network is a campus backbone where the departments run their own routers. One department is in two locations and wants to have connectivity for certain networks within it's department, but not make them known to the rest of the campus. I envision RIP-II Routing Domains as a replacement for this ability to unicast specific information to specific routers. Yes, some of this can be implemented with multiple IP subnets on a given wire. But I believe that would be harder for operations people to understand and support, and it still doesn't solve the problem of managing such a setup with SNMP. And the majority of forwarding engines that gated supports do not support multiple subnets on one wire. Jeff
- Comments on RIP-II protocol and MIB specification. Jeffrey C Honig
- Re: Comments on RIP-II protocol and MIB specifica… Fred Baker
- Re: Comments on RIP-II protocol and MIB specifica… Jeffrey C Honig
- Re: Comments on RIP-II protocol and MIB specifica… Fred Baker