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

[Jon Crowcroft]

the reaso nto use p2p to form the overlay is that p2p researchers have developed a bunch of
self organising scalable routing technologies which actually include research that has
moved on from the routing protocols used in IP -

most the ipv6 transisiton is still ip-centric-mindset which means it fails to scale:-)

of course one could deploy new routing protocols as part of ipv6/shipworm and other
transition tools, but i havent seen anyone talking about doing other than 
6-izing the existng raft of routing protocols, which doesnt buy s much in terms of
faster convergence, resilience to determiend attack, better defenses agains ddos, mobility,
multicast, or anything else we might dream up and actualyl deploy if we cut loose from the
path-vector+link state limited mindset of ip...


In message <F66A04C29AD9034A8205949AD0C901040194DC9F at win-msg-02.wingroup.windeploy.ntdev.microsoft.c
om>, "Christian Huitema" typed:

 >>> if the p2p service thus built (we might call it an InterNAT) has
 >>either
 >>> dynamic DNS update, or
 >>> uses ipv6 itself, then to provide global reachability is quite
 >>simple...
 >>
 >>Jon,
 >>
 >>Obviously, to merge multiple realms, you need an overlay; that overlay
 >>must provide a routing function; you can do the routing function either
 >>using explicit paths, a la UUCP, or global addresses and computed
 >>routing tables, just like IP. That is exactly what IP was designed for
 >>initially, e.g. merge an Ethernet realm and a packet radio realm. Do
 >>people still say Catenet? Now, how is that different from the
 >>combination of Shipworm/Teredo, 6to4 and other IPv6 transition tools?
 >>
 >>-- Christian Huitema

 cheers

   jon