Eric, >-----Original Message----- >From: Fleischman, Eric >Sent: Monday, November 24, 2008 5:14 PM >To: Templin, Fred L; Teco Boot; Robin Whittle; RRG >Subject: RE: [rrg] RACHH: the host-based solution - prepping the PR teamtosell it > >>From: Templin, Fred L >>Where I run into difficulties is when "edge-based" >>solutions are extended all the way into the core >>such that end sites are locked into a provider-aggregated "matrix". > >>Proper balance therefore depends on careful >>determination of where does "the edge" end and "the core" begin... > >I'm not tracking you here, Fred. Your recent I-D's (RANGER, VET, SEAL), when combined, naturally >suggest recursive deployments. In recursion, the concepts "edge" and "core" are local to a specific >recursive instance.
Yes, I think this is correct. The recursion can be applied from the core and continuing infinitely out to "the edges" however far removed from the core they may be. And, I agree with the observation that edge/core are local to their specific recursive instance. >For example, map-and-encaps solutions can be implemented in a recursive manner. In such a system, >edge and core are concepts which are local to a specific recursive instance. Given this, a provider- >aggregated matrix could only be visible if it occurred within the context of a specific recursive >instance -- it would be transparent to any other recursive instance, right? Yes; the aggregated matrix would only be visible within a particular recursive instance and, beyond that, it could (and possibly should) be all provider-independent addressing. Where I really have trouble is when some suggest that the aggregated matrix penetrates all the way into the core. That would be a total domination scenario for the ISPs, and in that lies madness... Fred [EMAIL PROTECTED] _______________________________________________ rrg mailing list [email protected] https://www.irtf.org/mailman/listinfo/rrg
