Hi, Scott Brim wrote:
- NAT is now architecture. The IETF needs to decide how much NAT it wants the Internet to have in the future. This strongly influences what we do in routing/addressing, because outlying, dwindling cases (either the NAT ones or the non-NAT ones) can be handled specially.
I suspect the IETF collectively knows that it would prefer exactly zero NAT deployment, but that it is currently stuck with it in an IPv4 universe. I further suspect that the IETF's preference for a future Internet would not include NAT, primarily for the technical reasons that a larger layer-3 address space renders NAT redundant, and that even the best-behaved NATs can cause problems for many transport protocols.
However, it seems that NAT is a technology people are unfortunately used to, that many people like, and that possibly some are uncomfortable without. Some recent efforts in the IETF (e.g., NAT66, HomeGate) at least acknowledge NAT is probably here to stay, while suggesting that standardised behaviour may ease future protocol development. In a sense, the "least evil" future of NAT. There's also the work on NAT traversal over existing non-standardised NAT44's, such as ICE.
I'm not convinced we arrive at an answer by throwing the question of dealing with NAT over the fence to the IETF. Might it be enough to assume for a future Internet that either: NAT deployment tends toward 0% as IPv6 deployment rises, or that future NATs adhere to certain well-defined behaviours for which we have a reasonable set of NAT traversal techniques to work around?
(Note in all of the above that I'm not involved in the IETF beyond passively watching a handful of mailing lists and remotely listening to some WG meetings.)
-S. _______________________________________________ rrg mailing list [email protected] http://www.irtf.org/mailman/listinfo/rrg
