Localization techniques for cluster-based data grid

In this paper, we present an efficient method for optimizing localities of data distribution when executing data parallel applications. The data to logical grid nodes mapping technique is employed to enhance the performance of parallel programs on cluster grid. Cluster grid is a typical computational grid environment consists of several clusters located in multiple campuses that are distributed globally over the Internet. Objective of the proposed technique is to reduce inter-cluster communication overheads and to speed the execution of data parallel programs in the underlying distributed cluster grid. The theoretical analysis and experimental results show improvement of communication costs and scalable of the proposed techniques on different hierarchical cluster grids.