In einer eMail vom 26.01.2009 19:35:37 Westeuropäische Normalzeit schreibt [email protected]:
> From: David Conrad <[email protected]> >> The question to ask is whether LISP is an appropriate solution to >> the problems discussed at the IAB's October, 2006 Routing and >> Addressing Workshop. >> ... >> Yet the LISP proponents still did not adequately address these concerns. > Perhaps the reasons LISP proponents haven't adequately addressed those > concerns is because they are not addressable with the underlying LISP > assumptions? When thinking about the routing, it's also important to remember that LISP, in the sense of 'a map-encaps scheme plus a mapping subsystem plus ancillary glue' is likely (almost certainly?) just a component, almost more of a piece of necessary substrate (to give us a way out of the current paint corner), than a full-blown new routing approach. (I've referred before to how LISP reminded me of simply the deployment plan for a new routing architecture I worked on.) Yes, LISP may do some routing-like things (e.g. support multi-homing, and provide a certain amount of provider independence), but there's a limit to what you can do with simply separating location and identity. Right. At the end of the day, we're still working with BGP, and if you want to do more with _routing_ than BGP allows (e.g. control the paths for traffic aggregates), then you need something _else_ new. Formerly I also thought, overcoming the DV algorithm required to replace BGP. But meanwhile I see it differently: The fact that BGP is all present is a big help and not a big obstacle. Routing without disseminating any user r eachability prefixes is certainly something_else_new. But it can be enabled by enhancing BGP ! And deploying that new thing would almost certainly start with.... deploying something that looks a lot like LISP. Add that to the fact that LISP gives you a decent bang/buck _in and of itself_ (multi-homing, etc), and you soon decide to focus on getting LISP done (right, with hooks for the future), and worry about the rest later... LISP will never get IP forwarding on the Moore's law applicability track. That's for sure :-( Heiner Noel _______________________________________________ rrg mailing list [email protected] http://www.irtf.org/mailman/listinfo/rrg
_______________________________________________ rrg mailing list [email protected] http://www.irtf.org/mailman/listinfo/rrg
