A High-Performance Intelligent Integrated Data Services System in Data Grid

Abstract

It is very complex and of low efficiency for grid users to access heterogeneous diverse data resources distributed over the whole wide area network. The integrated data services can implement more wide range of transparent access to data resources in Data Grid. One of the important building blocks for any Data Grid is the server which implements data services. In this paper, we describe HPIIDS which implements integrated data services and achieves more wide range of transparencies. In order to improve its performances, we emphasis on the server called DSB which provides integrated data services. The DSB bases on cluster and agent technologies. Agent-based DSB can cleverly prefetch required data, replicate them among DSBs. Multiple DSBs are formed cluster HPIIDS. Cluster-based HPIIDS implements a single entry, provides a virtual integrated storage system for users, hides distributed heterogeneous low-level data resources, and insures load balance of each server. HPIIDS architecture supporting these various scenarios, are also described.

This paper is supported by the National Natural Science Foundation of China under the Grant No. 60203016 and No. 69933030, and the National Hi-Tech R&D 863 Program of China under the Grant No. 2002AA131010.