Hello, We were among the first one to run large scale BitTorrent experiments on clusters instead of classical platforms like PlanetLab. Our argument in favor of a cluster is the scale (we ran up to 10 000 simultaneous real BitTorrent peers, but is is possible to run much more) and the reproducibility.
However, we received criticism arguing that PlanetLab experiments are more realistic. PlanetLab offers a large variety of delays and loss rates but due to virtualization it fails to be realistic. For this reason, we performed a study to evaluate the impact of delays and losses on BitTorrent experiments. We found that realistic BitTorrent experiments can be safely run on clusters. We believe that this work is relevant for this community. Here is the reference to this work that has been accepted in the P2P'10 conference:http://hal.inria.fr/inria-00494410/en/ Regards, Ashwin _______________________________________________ p2p-hackers mailing list p2p-hackers@lists.zooko.com http://lists.zooko.com/mailman/listinfo/p2p-hackers