Data grid performance analysis through study of replication and storage infrastructure parameters

Running data grid applications such as high energy nuclear physics (HENP) and weather modelling experiments involves working with huge data sets possibly of hundreds of Terabytes to Petabytes in size often kept over wide area networks. Data replication is a useful technique for reducing latency across communication networks over which the source data are accessed. As a starting point towards developing a multifaceted optimisation solution for data grids, this paper considers the effect of replication and storage parameter settings on data grid performance. The simulation results we obtained suggest that replication at local (Tier2) nodes has significant impact on data grid performance while cache settings at remote (Tier 1) node result in minimal performance improvement.