On Mar 12, 2013, at 11:07 AM, John Curran <[email protected]> wrote: > On Tue, 05 Mar 2013 12:40:43 -0800, Joel M. Halpern wrote: >> Is there a way to structure things with unified, automated, assignment now, >> but the ability to split along this and other dimensions later? (We don't >> even need to spell out the split, just be confident we can do it if we want >> to. > > This particular approach looks to be the worst of all worlds, in that you'll > have > a deployed solution which requires significant effort and coordination to > handle > the split, and no clear reason to actually make the changes since everyone is > just fine with interim state (hmm...kinda like the first 15 years of IPng ;-)
I personally continue to believe that doing a registry/registrar split for EIDs at this point of time would be making things _way_ too complicated for the level of deployment LISP has enjoyed to date. It's sort of like trying to jump from RFC 883 DNS to ICANN without going through the RFC 1034/1035 (et al.) and SRI-NIC->DDN NIC->GSI->NetSol->VeriSign->IAHC->White/Green Paper->NewCo steps first. I also feel even going down the RIR route is far too heavyweight, even ignoring the implications of regionalization. Saddling LISP EID allocation with unnecessary bureaucracy, complexity, and regional politics will simply discourage/delay experimentation and/or adoption. At this stage of the game, I believe a simple, unified and centralized registry with a straightforward non-discriminatory allocation policy is quite sufficient. If folks are uncomfortable with the scalability of that model, adding text that describes an evolutionary plan (that could include a registry/registrar split) should the centralized registry be unable to cope with demand might make sense, but I don't think we need to do much more that document the plan. Regards, -drc _______________________________________________ lisp mailing list [email protected] https://www.ietf.org/mailman/listinfo/lisp
