A Hybrid Architecture for Content Consistency and Peer Synchronization in Cooperative P2P Environments

Abstract

Peer-to-Peer architectures for content and knowledge management foster the creation of communities of workers in which effective knowledge and information sharing takes place. In such communities, workers have similar capabilities in providing other workers with data and/or services and are autonomous in managing their own knowledge objects. Since objects are typically shared among a set of workers, problems regarding concurrent access to content, content consistency and synchronization of peers arise. This paper describes a hybrid architecture for the management of data consistency and peer synchronization. The designed framework combines centralized, yet dynamic, mechanisms for metadata management and peer-to-peer mechanisms for data transfer. The paper reports on the use of these mechanisms in K-link+ a P2P collaborative platform, developed at the GridLab of the University of Calabria, for distributed knowledge management. An analytical study founded on queue network theory confirms the efficiency of the presented approach.