DefinePK

Abstract

Data Grid is a geographical-based distributed environment that deals with the extensive data-oriented applications. Data replication is used for reducing the data access latency and managing effective processing of huge data in the distributed environment. This research work presents a hybridized Modified Dynamic Hierarchical Replication Algorithm (MDHRA) and Efficient Replica Consistency Model (ERCM) for the data grid environment. The location of the best file replica is chosen based on the data transfer time, storage access latency, replica requests in the queue and distance between the nodes in the data grid environment. The ERCM model requires minimum execution time for the reading and writing operations that ensures high data availability. The main contributions of the ERCM approach are optimal allocation of replica and maintenance of the replica consistency. The experimental analysis shows that the combined approach requires minimum Effective Network Usage (ENU), storage usage, replication frequency, number of communications and maximum hit ratio than the existing replication techniques.