Fifth International Conference on Grid and Cooperative Computing Workshops
A Data Transfer Scheme of Grid Workflow Based on Weighted Directed Graph
Hunan, China
October 21-October 23
ISBN: 0-7695-2695-0
Song Wu, Huazhong University of Science & Technology, China
Hai Jin, Huazhong University of Science & Technology, China
Kang He, Huazhong University of Science & Technology, China
In a grid workflow, data transfer scheme plays a key role on its performance. In most grid projects, data is transferred between source and destination directly in the traditional way, without considering the status of network. In this paper, we present a data transfer scheme of grid workflow based on weighted directed graph. The scheme selects data transfer model according to the job's data processing rate. Based on the different bandwidths between grid nodes, it builds a weighted directed graph to select a best transfer path. Simulation results show that the scheme can effectively improve the performance of grid workflows.
Citation:
Song Wu, Hai Jin, Kang He, Zongwei Luo, "A Data Transfer Scheme of Grid Workflow Based on Weighted Directed Graph," gccw, pp.491-495, Fifth International Conference on Grid and Cooperative Computing Workshops, 2006