[p2p-hackers] Work on NAT-friendly DHTs?

[Matthew Kaufman]

Philip Matthews:
> I am wondering if anyone is aware of work on NAT-friendly DHT 
> algorithms?
> In other words, a DHT algorithm that works with the peers are 
> located behind NATs.

We're working on that here at amicima. One of the functions provided by our
MFPGroup layer is DHT-style routing, and it is based on our existing (and
open-source) MFPNet work that provides NAT traversal.

>  From what I have seen. most DHT algorithms assume that any 
> peer is directly reachable from any other peer...

Most of the first round of DHT algorithms are research projects that don't
solve a number of real-world problems. This is starting to change, and I
think you'll be seeing more "usable in real products" algorithms and
implementations in the near future. We, for instance, are tackling full
participation by behind-NAT peers, cryptographic control of membership, high
churn rate, and application-level multicast in our project. Which is
probably why it hasn't shipped quite yet :)

Matthew Kaufman
matthew at matthew.at
http://www.amicima.com