Measurement, modeling, and analysis of a peer-to-peer file-sharing workload

Abstract

Peer-to-peer (P2P) file sharing accounts for an astonishing volume of current Internet traffic. This paper probes deeply into modern P2P file sharing systems and the forces that drive them. By doing so, we seek to increase our understanding of P2P file sharing workloads and their implications for future multimedia workloads. Our research uses a three-tiered approach. First, we analyze a 200-day trace of over 20 terabytes of Kazaa P2P traffic collected at the University of Washington. Second, we develop a model of multimedia workloads that lets us isolate, vary, and explore the impact of key system parameters. Our model, which we parameterize with statistics from our trace, lets us confirm various hypotheses about file-sharing behavior observed in the trace. Third, we explore the potential impact of locality-awareness in Kazaa.Our results reveal dramatic differences between P2P file sharing and Web traffic. For example, we show how the immutability of Kazaa's multimedia objects leads clients to fetch objects at most once; in contrast, a World-Wide Web client may fetch a popular page (e.g., CNN or Google) thousands of times. Moreover, we demonstrate that: (1) this "fetch-at-most-once" behavior causes the Kazaa popularity distribution to deviate substantially from Zipf curves we see for the Web, and (2) this deviation has significant implications for the performance of multimedia file-sharing systems. Unlike the Web, whose workload is driven by document change, we demonstrate that clients' fetch-at-most-once behavior, the creation of new objects, and the addition of new clients to the system are the primary forces that drive multimedia workloads such as Kazaa. We also show that there is substantial untapped locality in the Kazaa workload. Finally, we quantify the potential bandwidth savings that locality-aware P2P file-sharing architectures would achieve.

Evaluation of P2P multimedia clustering techniques in JXTA-Overlay

“…Gummadi et al [8] consider clustering by file type. The authors, after analyzing large traces of P2P traffic, proposed a model of multimedia workloads aiming to explore the impact of key system parameters.…”

“…Clustering techniques have attracted the attention of researchers from the P2P community due to their usefulness in developing efficient searching in P2P file sharing systems. Presently, three types of clustering techniques have been studied for P2P networks: locality-awareness clustering [9], clustering by file type [8] and clustering by content of textual files [10,18,22,27]. However, no studies have been carried out for clustering of multimedia files by content in P2P networks.…”

Characterizing files in the modern Gnutella network

“…(2) overlay topology structure [4,27,28,5], (3) query traffic [6], (4) data traffic [29,25,30], and (5) shared files [7,8]. We are aware of only two other studies that focus on the characteristics of files shared by users.…”

“…Gummadi et al [29] analyzed a 200-day trace of Kazaa traffic collected at the University of Washington, demonstrating that file transfers in Kazaa do not follow a Zipf distribution and argued that this difference is due to the "fetch-at-most-once" nature of downloads in file-sharing applications. Another analysis of Kazaa traffic was conducted by Leibowitz et al [25] at a large Israeli ISP.…”

Optimizing self-organizing overlay network using evolutionary approach

“…We simulate flooding search used in a Gnutella network by conducting the Breath First Search algorithm from a specific node. Basic settings of the workloads used in this simulation follow a 200-day trace of KaZaA P2P traffic collected at University of Washington [19].…”