[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Enhanced SIIT

This is something potentially useful that I think deserves further review - simply because the approach appears to provide the ability to use a stateless transform of the protocol header in the network.
It its current formulation it requires host protocol stack surgery, but it you look past the precise details and think about the essential differences between NATs with explicit bindings and held state and algorithmic transforms that do no require kept state then there is something here that could be a useful starting point. 

So I'm of the view that it merits a little further attention.

Geoff

Iljitsch van Beijnum wrote:
After a half serious post to the IETF list I decided to write a quick draft about the following: 

Abstract

  This document describes an extension to the mechanism outlined in RFC
  2765 that allows IPv4 hosts to communicate with IPv6 hosts through a
  protocol translation device with full IPv6 compatibility.

  For this purpose, a new header is inserted between the IP header and
  the payload protocol such as TCP and UDP. The new header contains the
  bits truncated from the IPv6 address when the IPv6 address is
  translated into an IPv4 address in the 240.0.0.0/4 (class E) range.

http://www.muada.com/drafts/draft-van-beijnum-v6ops-esiit-00.txt

(I have no patience with the IETF tools right now, will retry that later.)
The idea is that hosts with IPv4 connectivity get to talk to IPv6 hosts without anything NAT-like getting in the way. However, it does require some changes on the IPv4 side, but these can be deployed incrementally. The assumption is that these changes would be easier to deploy than a full IPv6 transition.
I would really like some feedback on whether this is something useful that deserves more attention or not.