A Resource Broker with Cross Grid Information Services on Computational Multi-grid Environments

Grid computing is now in widespread use, which integrates geographical computing resources across multiple virtual organizations to achieve high performance computing. In this paper, we present a new grid architecture, which integrates multiple computational grids, named Multi-Grid, from different virtual organizations. The proposed architecture could solve the large-scale computational problem. Due to gathering information across different grids has the imprecise problem, the accuracy of information probably affects the computational efficiency. Thus, this paper proposes a cross-grid information service approach to gather resource information. Furthermore, a gathering algorithm, which could balance the query frequency and the resource information accuracy, is present to reduce the load of information servers. Finally, we implement a resource broker with cross-grid information services on multi-grid computational environments.