Miscellaneous notes related to Master Thesis project 1. DHTs Discussion with the author of the Kademlia p2p system (on which eDonkey2000 and Overnet are based on): - Start - > Are there any studies/articles available about these kind of features of DHTs ? P2P systems are a new phenomenon, so not many are available. Look at the proceedings of the 1st International Workshop on P2P systems and see the papers and other papers they reference. That's pretty much all that is known about P2P systems. You can find the site by looking for "IPTPS'02" on Google. PEtar > So in theory, it should scale to 1000's millions of users (billions), right ? Yes. The only constraint is that the routing table needs to be maintained in a timely manner. The routing table has size proportional to log n, where n is the number of peers in the system. So as long as the Internet is fast enough to maintain this size routing table, we are good. People think that this is the case. Petar > Are there any weak points in your systems from the scalability/efficieny point > of view ? There's the problem what to do when one word (e.e. "mp3") is very popular. So the question is where do you store it and how do you deal with it. But this is not a problem of Kademlia, rather it is a general problem of DHTs (Distributed Hash Tables) - End -