Tunneling legacy routers a la LISP presumes global ipv4 address uniqueness, i.e. won't work as soon as that presumption doesn't hold anymore.
This is a major difference with TARA. Because TARA routers do not disseminate user reachability, each TARA router which replaces a legacy router would free a bunch of ipv4 addresses from the necessity to be globally unique. LISP reminds me of MADCAP. As soon as MADCAP was accomplished it was OUT. And the above is one of the reasons why I expect LISP to share with it the same destiny. Heiner -----Original Message----- From: Templin, Fred L <fred.l.temp...@boeing.com> To: rrg <rrg@irtf.org> Sent: Thu, Dec 6, 2012 1:03 am Subject: [rrg] Where IRON fits in Good to see some activity on this list - I think? A couple of posts have mentioned IRON, and that work did indeed originate from this group. However, I have come to the realization that IRON is really an interior routing solution and not an Internetworking solution. In particular, IRON does not use tunneling between the routers of separate sites across the DFZ the way LISP does; it only uses tunneling *within* a site, even if that "site" is distributed across the global Internet. Where IRON adds value is that it manages mobility and multi-homing within the site without exposing de-aggregated prefixes to BGP. The IRON relays can therefore advertise only a few short prefixes into the BGP instead of lots of long prefixes. I therefore think IRON is complimentary to either of LISP or ILNP. It's just that IRON does well for "sites" that are geographically distributed and/or have lots of mobility internally. Thanks - Fred fred.l.temp...@boeing.com _______________________________________________ rrg mailing list rrg@irtf.org http://www.irtf.org/mailman/listinfo/rrg
_______________________________________________ rrg mailing list rrg@irtf.org http://www.irtf.org/mailman/listinfo/rrg
