A survey on data replication strategies in a Data Grid environment

Article type: Research Article

Authors: Naseera, Shaik

Affiliations: School of Computing Science and Engineering, VIT University, Vellore, India. E-mail: [email protected]

Abstract: In a Data Grid Environment, data is often replicated to several sites in order to improve data availability, fault tolerance and faster data access to user requests. Data Grids deals with data intensive applications, which require data to be replicated at multiple locations for the improvement in the overall performance of the application. Data replication must address some of the issues like when to create/delete a replica, how many copies of replicas to create, which files are to be replicated and selection of location for data replication. Several data replication strategies have been proposed in the recent years for Grid computing environment. In this paper, we present a thorough survey on the issues and challenges involved in data replication focusing on the contemporary issues like replica consistency, synchronization and maintenance. A general overview has been given on data replication, describing when replicas can be created/modified and which replica to be selected for use, maintaining optimal number of replicas, maintaining replicas consistency, limitations and their future enhancements.

Keywords: Grid computing, Data Grid, data replication, replica creation, replica maintenance, replica synchronization, replica consistency, optimization, survey

DOI: 10.3233/MGS-160253

Journal: Multiagent and Grid Systems, vol. 12, no. 4, pp. 253-269, 2016