A Grid Resource Broker with Dynamic Loading Prediction Scheduling Algorithm in Grid Computing Environment

摘要

In a Grid Computing environment, there are various important issues, including information security, resource management, routing, fault tolerance, and so on. Among these issues, the job scheduling is a major problem since it is a fundamental and crucial step in achieving the high performance. Scheduling in a Grid environment can be regarded as an extension to the scheduling problem on local parallel systems. The NCHC research team developed a resource broker with proposed scheduling algorithm. The scheduling algorithm used the adaptive resource allocation and dynamic loading prediction functions to improve the performance of dynamic scheduling. The resource broker is based on the framework of GridWay which is a meta-scheduler for Grid, so the proposed resource broker can be implemented in real Grid Computing environment. Furthermore, this algorithm also has been considered several properties of the Grid Computing environment, including heterogeneous, dynamic adaptation and the dependent relationship of jobs.