A SIMPLEX GRID COMPUTING METHODOLOGY FOR MONITORING RESOURCES

[摘要] Controlling large scale distributed system is very difficult for the reliable operations of a system. This is because of two reasons. One, wide fluctuations in the availability of idle processor cycles. Second, communication latencies over multiple administrative domains. So, the quality of service in executing grid applications becomes a challenge. With such fluctuations there is no certainty when a grid task will complete its execution time and therefore become unpredictable. To overcome this problem, grid computing technology is used. More specifically a data grid is used to aggregate the unused storage space into a much large virtual data store. This is further configured to achieve improved performance and reliability. Thus it has the possibility of substantially increasing the efficiency of resource usage. A three-tier architecture of Distributed Computing Grid Model (DCGM) is brought forward. It acts as the infrastructure of multi-site scheduling environment in which Distributed scheduling server is utilized. First Come First Serve (FCFS) algorithm is implemented. Thus the applications of splitting the numbers depending upon the CPU idle rate is established and thereby reducing the execution time.

[关键词] Distributed system;Grid;Applications;Grid Computing;Virtual data store;Autonomic Computing