On Mon, May 14, 2012 at 9:40 AM, Serguei Osokine <serguei.osok...@efi.com> wrote: > Anyway, it sounds like whatever these guys are claiming to their > investors, they might have a bit of a problem killing BitTorrent, > unless they are also doing something else that would be unrelated > to DHT poisoning.
Well it sounds so far like it's a 1-2 punch of: 1) Flood the trackers with fake peers 2) Poison the DHT with fake nodes (who give out more fake peers) 3) (??) Have fake peers give out more fake peers via PEX I could imagine that would be an effective strategy on a torrent-by-torrent basis. But it would also be enormously expensive. The real trick I think is (2), as trackerless torrents would be totally defeated if you could take over the right nodes in the DHT. How many nodes ultimately control a given torrent? Does it all come down to a single node, or is there some kind of load-balancing across multiple nodes? (I imagine there's a lot of "failover", but not sure how many DHT nodes actively manage a single torrent.) I wonder if the startup might be using some sort of cloud SHA1/public-key cracking service to come up with node IDs or even fake torrent files. Sounds fun! -david _______________________________________________ p2p-hackers mailing list p2p-hackers@lists.zooko.com http://lists.zooko.com/mailman/listinfo/p2p-hackers
