> From: Scott Brim <[EMAIL PROTECTED]>
>> Excerpts from Noel Chiappa on Thu, Dec 04, 2008 11:05:56AM -0500:
>> we are, in effect, building a second packet-switching system on top of
>> the first, but [somewhat] separate from it
> As argued elsewhere ... in order for it to be a different "packet
> switching system", it would have to have different addressing, routing,
> and switching.
Briefly (to avoid recapitulating all that's been said elsewhere) - I think
this view that it's all part of the same thing is an illusion, because in the
initial stages of LISP, it _looks_ like it uses the 'same' addresses, etc.
Imagine that we replaced the inter-xTR protocol with something completely
different (maybe raw MPLS packets, using, say, PNNI for routing, and with a a
new address space, etc). You'd still have all the _exact_ same issues that we
do with IPv4-based designs (which use the legacy Internet as a giant NBMA
network) - i.e. picking an exit point, making sure it was up and reachable,
etc.
I think in that circumstance, it's clear that my claim that this is 'another
packet switching layer' would definitely be accurate, right? And those issues,
which we would also face with the IPv4-based design, are ones one typically
faces in designing a packet-switching system, no?
Yes, I know the picture is more complex (e.g. we have partial visibilitiy into
the NBMA box, e.g. via a BGP feed which can tell us reachability of
designation ETRs). But I think it's a more useful model - i.e. one in which
the problems and challenges are clearer and easier to think about - if one
sees it as a new layer, and not an extension to an existing one.
Noel
_______________________________________________
rrg mailing list
[email protected]
https://www.irtf.org/mailman/listinfo/rrg
