Re: GSE IDs [Re: IETF multihoming powder: just add IPv6 and stir]

[Brian E Carpenter <brian@hursley.ibm.com>]

Iljitsch van Beijnum wrote:
> ... NAT shows that it is
> possible to build middleboxes that keep lots of state.

and the failure modes associated with NATs show why
this is a really bad idea. But even NATs keep that state
reasonably local; it doesn't have to be known by magic
to an anti-NAT at the other end.

Iljitsch, the big hole in your -00 draft is that is doesn't
at all discuss the magic needed to distribute mapping state
and keep it fresh. As I think I already said, this is why we
never pursued the map-and-encap proposal some years ago- it's
not a side issue, it's *the* issue in this class of solutions,
and it is the fundamental difference from 8+8/GSE.

I believe MHAP does discuss this question. And as Jim said, there
are stateless solutions, but that means adding bits to the packet.

  Brian