> -----Original Message----- > From: [email protected] [mailto:[email protected]] On > Behalf Of Scott Brim > Sent: 07 May 2009 12:43 > To: Toni Stoev > Cc: IRTF RRG > Subject: Re: [rrg] Inter/intra-domain routing separation > > Toni Stoev allegedly wrote on 05 05 2009 5:35 AM: > > Intra-domain routing can be considered as a general solution. This > > general solution is the provision of reachability throughout an > > autonomous system. Node locators can be considered intra-domain > > locators. Every locator shall have default association with its > > containing autonomous system in order to be universally > recognizable. > > First, see http://tools.ietf.org/html/rfc1955 > > Second, IMHO AS#s are the wrong granularity. > > - a prefix can be reachable via more than one AS > > - there can be more than one prefix within an AS > > - it should be possible to route different prefixes, and > subprefixes, > via different paths for traffic engineering purposes.
Why should we do that at the level of prefix - would it not be nicer to route different traffic via different paths to load balance based on load - who really wants to prefix balance. Also, a computer can be reached by more than one prefix as well as more than one AS If we assume we have an end point identifier then would AS be the wrong level of granularity for the top level of a routing heirarchy? Seems to me that the prefix level is too detailed for us to handle well so we need to make the top level less fine grained? Louise > > Scott > _______________________________________________ > rrg mailing list > [email protected] > http://www.irtf.org/mailman/listinfo/rrg > _______________________________________________ rrg mailing list [email protected] http://www.irtf.org/mailman/listinfo/rrg
