[e2e] using p2p overlays to overcome recursive NATs/realms

[Jon Crowcroft]

In message <Pine.SOL.4.43.0202081232310.4609-100000 at phaestos.ee.surrey.ac.uk>, Lloyd Wood typed:

 >>> where one can infer the infernal internal algorithm used by a nat,
 >>> use it.
 
 >>the same basic principle that NAT ALGs use to handle the infernal
 >>layer-breaking assumptions of various protocols. And look how well
 >>that worked... better to be simple and compelling than fight
 >>complexity with added complexity.

right, but:
see also

http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-nat-scenarios-00.txt
http://www.ietf.org/internet-drafts/draft-ietf-sip-nat-01.txt
http://www.ietf.org/internet-drafts/draft-rosenberg-midcom-stun-00.txt
http://www.ietf.org/internet-drafts/draft-rosenberg-midcom-turn-00.txt
http://www.ietf.org/internet-drafts/draft-ietf-mmusic-sdp4nat-00.txt
http://www.ietf.org/internet-drafts/draft-davies-fw-nat-traversal-01.txt
http://www.ietf.org/internet-drafts/draft-sen-midcom-fw-nat-00.txt



 
 >>> if the p2p service thus built (we might call it an InterNAT)
 >>bt,dt.
 
 >>"The InterNAT: Policy Implications of the Internet Architecture
 >> Debate", by Hans Kruse, William Yurcik and Lawrence Lessig,
 >>Proceedings of the Telecommunications Policy Research Conference
 >>(TPRC), Virginia, September 2000.

ok - thanks...
 >>
 >>http://www.sosresearch.org/publications/tprc00.PDF
 >>http://www.sosresearch.org/publications/
 
 >>an idea whose time has unfortunately come.


the reason i mentioend p2p is that some of the good p2p systems have nice workarounds
for self organising - thus they may route around the damage that NAT causes to the
internet quite well - the goal is to revisit the original design objective of IP that
seems to be lost on recent folk...robustness:)

some of the current straight IP topology is potentially a target for attack (c.f.
Réka Albert, Hawoong Jeong and Albert-László Barabási, "Error and
attack tolerance of complex networks," Science, vol.
406, pp. 378--382, Jul. 2000.)
and some of the fingerprint routing in chord and other systems could mitigate against
this if migrated into the infrastructure eventually, and we should be experiemtning with
it in overlays right now, and a NAT Overlay Tapestry (NOT, (tm)) would give folk a
incentive maybe to deploy



 cheers

   jon