Performance Modelling and Evaluation of Event-Condition-Action Rules on RDF in P2P networks

This paper studies the performance and scalability aspects of processing Event-ConditionAction (ECA) rules on RDF metadata in peer-to-peer (P2P) environments. This work has been motivated by the increasing use of RDF in distributed web-based applications that require timely notification and propagation of metadata changes between peers, for which ECA rules are a possible candidate technology. We develop a performance model for analysing the behaviour of a rule processing system for ECA rules on RDF metadata in P2P environments. We use as the main performance criterion the time required to complete all rule processing resulting from a top-level update submitted to one of the peers in the network. We examine how this time varies with the network topology, number of peers, number of rules, and degree of metadata replication between peers. We describe a simulation of the system, and the analytical results are compared with analogous experiments conducted with the simulation, which show good agreement. Our overall conclusions are that if a HyperCup topology is used for interconnecting the superpeers, then ECA rule processing on RDF metadata in P2P environments shows good scalability, pointing to its practical usefulness as a technology for real applications.