Mapping the Gnutella Network

Peer-to-peer systems have emerged as a significant social and technical phenomenon, and they are likely to gain popularity as low-cost individual computing and storage resources become more widely available and network connectivity increases. Unlike traditional distributed systems, P2P networks aggregate large numbers of computers that join and leave the network frequently and that might not have permanent network (IP) addresses. This new breed of systems creates application-level virtual networks with their own routing mechanisms that allow individual computers to share information and resources directly, without dedicated servers. The topology of these overlay networks and the routing mechanisms used have a significant impact on application properties such as performance, reliability, scalability, and, in some cases, anonymity. The topology also determines the communication costs associated with running the P2P application, both at individual hosts and in the aggregate. Note that the decentralized nature of pure P2P systems means that these are emergent properties, determined by local decisions made by individual resources based only on local information: We are dealing with a self-organized network of independent entities. To explore these issues, we studied the topology and protocols of the public Gnutella network. Its substantial existing user base and open architecture made it a good large-scale, if uncontrolled, testbed. We captured the network’s topology, generated traffic, and dynamic behavior to determine its connectivity structure and how well (if at all) Gnutella’s overlay network topology maps to the physical Internet infrastructure. Our analysis of the network allowed us to evaluate costs and benefits of the P2P approach and to investigate possible improvements that would allow better scaling and increased reliability in Gnutella and similar networks.