Challenges in P2P Computing

Abstract

Peer-to-peer (P2P) is an emerging model aiming to further utilize Internet information and resources, complementing the available client-server services. P2P has emerged as a promising paradigm for developing large-scale distributed systems due to its many unique features and its potential in future applications. P2P systems are popular because of their adaptation, self-organization, load-balancing, and highly availability. However, P2P systems also present many challenges that are currently obstacles to their widespread acceptance and usage, such as efficiency, security, and performance guarantees. For example, studies have shown that P2P traffic contributes the largest portion of the Internet traffic based on the measurements on some popular P2P systems, such as FastTrack (including KaZaA and Grokster), Gnutella, and Direct Connect. Even given that 95% of any two nodes are less than 7 hops away and the message time-to-live (TTL=7) is preponderantly used, the flooding-based routing algorithm generates 330 TB/month in a Gnutella network with only 50,000 nodes. In reality, there are millions of active P2P users at any given time. Our study has shown that the mechanism of a peer randomly choosing logical neighbors without any knowledge about the underlying physical topology causes topology mismatch between the P2P logical overlay network and the physical underlying network. A large portion of the heavy P2P traffic is caused by inefficient overlay topology and the blind flooding. Security and anonymity are other concerns in P2P systems. This talk will address the above issues as well as other potential applications of P2P computing and mobile P2P systems.