-----BEGIN PGP SIGNED MESSAGE----- Per Gregers Bilse <bilse@EU.net> writes:
What would happen if the customer used private address space and NAT & ALGs to hook this into the two providers' CIDR space?
I think that this is an excellent idea. In fact, Ed Kern at Digex is working with OFRV on making this real, so that other folks can do so when the NAT code is available in IOS. He might have something to say about your note.
Has anybody noticed that the net global routing table has bumped 40k a couple of times recently?
Yes. Fun fun fun. Has anyone else seen SSE microcode size get too big as a result? Ping. More than 2/3 of the way to the point where 7000s+SSPs will not be usable with traffic towards a fair number of destinations, in routers that carry full routing.
Would the registries have problems with this approach?
Would it matter if they did? :-) No, seriously, any concerns about the cases where twice the PA address space will be used should be releived by the reduction in demand (or at least in absolute need) for PI address space. In short, if it works well enough to be a general solution for customer migratability without requiring migratable address allocations, it's obviously a net win for the registries too.
Another issue is that the I-must-be-multihomed-to-different-providers mantra frequently isn't anything else than a funny idea in somebody's head.
Well, I agree with you about market misperception. However, this doesn't seem to be a very popular thing to say (flame wars ahead), and in part this is because few people have been saying it in public, even when they actually agree. Sean. -----BEGIN PGP SIGNATURE----- Version: 2.6.2 Comment: PGP Public Key in ftp://ftp.sprintlink.net/engineer/smd/pgpkey iQCVAwUBMhZE7kSWYarrFs6xAQEvegQAjTHWfPO1Kbeuc6TYgFJ16EV6KjTq22AM tcW/VGP7NJ3DDuySlDgHqa2180NKrDlBGbU9kyK9nZJB67eWtvigEJ+V4T+D33aP vlJXNItzbQR7X9hnbAHr386VGDFjdm+CrnvC6fPGUZm754B2AbwCI/4qVJ9PZy2s wU78IyMudiM= =TWtU -----END PGP SIGNATURE-----