Performance analysis of a locality-aware BitTorrent protocol in enterprise networks

Nowadays, distributions of large volumes of data over enterprise TCP/IP networks bring problems such as high network and server utilizations, long periods for completion, and greater sensitivity to flaws in network infrastructure. These problems can be reduced with the use of Peer-to-Peer (P2P) networks, but the standard BitTorrent protocol does not perform well in some scenarios such as the one related to enterprise networks. This work proposes a modification in the BitTorrent protocol that uses locality-awareness to reduce traffic on long distance links of enterprise networks. In this modification, the tracker identifies the peer IP address requesting the list of IP addresses of the swarm and sends only those addresses that belong to the same LAN and to the original seeder. The performance of the proposal has been assessed by simulations. In typical enterprise scenarios, results show that the change is able to reduce bandwidth consumption and time of downloads compared with standard BitTorrent, and gives greater robustness to the distribution in case of failure of WAN links. Results also show that in more complex network environments, with many clients and where the bandwidth restriction on long distance links causes congestion and packet drops, the proposed modification consistently improves distribution performance.