Abstract

Grids can be classified as computational grids, access grids and data grids. Computational grids address applications that deal with complex and time intensive computational problems, usually on relatively small datasets. Access grids focus on group-to-group communication. Whereas data grids address the needs of applications that deal with the evaluation and mining of large amounts of data in the terabyte and petabyte range. While SOR-CER is a federated computation grid environment, a complementing data grid service called Replica Provider is introduced. The newly developed data grid service is used and at the same time as the already existing SORCER compute grid is leveraged for an increased functionality. SORCER service-oriented programs with replication services now has the capability of running data grid applications. It enables better access to such databases and increases reliability by replicating them at multiple locations. A federated grid environment with replication service is presented. It enables large number of files to be replicated on multiple nodes over different heterogeneous computation platforms and provides generic service providers for fast, up-to-date, and reliable access to file storage. 1

Citations