Optimizing Replica Creation using Agents in Data Grids

摘要

Data Grid is emerging as a key technology to support data intensive applications. Grids rely on data replication to enhance performance of the system. Maintaining replicas is the primary concern of replica management in grid environment. This paper addresses the issue of replica creation which in turn enhances the accessibility of updated replica in grid environment. A strategy for Replicas Creation based on Popularity (RCP) has been developed which dynamically creates the replicas in distributed grid environment. An intelligent agent based approach is used which helps in optimizing the number of replicas based on the access frequency of a file. The proposed RCP strategy is implemented on Optorsim simulator and the results are compared with existing approaches to judge its effectiveness in data grids. The simulation results show significant benefits of proposed algorithm in terms of average execution time, number of replica and effective network usage.