Clustering in Peer-to-Peer File Sharing Workloads

Abstract

Peer-to-peer file sharing systems now generate a significant portion of Internet traffic. A good understanding of their workloads is crucial in order to improve their scalability, robustness and performance. Previous measurement studies on Kazaa and Gnutella were based on monitoring peer requests, and mostly concerned with peer and file availability and network traffic. In this paper, we take different measurements: instead of passively recording requests, we actively probe peers to get their cache contents information. This provides us with a map of contents, that we use to evaluate the degree of clustering in the system 1 , and that could be exploited to improve significantly the search process.

Revealing intricate properties of communities in the bipartite structure of online social networks

“…For computational reason, we did not use the whole network and rather focused on a large connected component 1 . This sub-network involves 3 millions articles and almost 500,000 subcategories.…”

“…This approach have proven to be useful to identify non trivial properties of the structure of networks in very different contexts. We can cite for instance computer networks (like the Internet, peer-to-peer systems, the web) [1], [2], biological networks (protein-protein interaction networks, metabolic processes) [3], [4], social networks (friendship networks, co-publication networks) [3], [5], legal networks [6], linguistics [7], economy [8], etc.…”

Correlating Topology and Path Characteristics of Overlay Networks and the Internet

“…A large-scale study of the eDonkey network is presented in [10]. The authors' focus is on clustering the peers' preferences; the peers' Internet connectivity is not analyzed, and from the peers' locations only the countries are detailed.…”

Extending JXTA Protocols for P2P File Sharing Systems

“…A as a matter of fact, there are still many of issues to understand and better address in P2P file sharing domain. Thus, deciding which protocols to use [6] (Freenet, Gnutella, and Napster use different protocols), how to reduce the traffic generated due to file sharing [3], [4], what schemes to use for efficient indexing/searching files within a P2P file systems [2] are such research questions, to name a few. On the other hand, there are issues related to the design of P2P systems for file sharing that benefit from the decentralized nature of P2P systems yet facilitate efficient file sharing among peers and P2P file sharing applications.…”