Job schedule model based on grid environment

Grid computing architecture was defined to be a complete physical layer. The functions of information systems based on grid architecture are resources sharing, collaborative processing, etc. System performance is calculated from resource usages of all grid nodes. If we have load-balancing in all grid nodes, we will have good performance in the system. All the grid nodes are divided to supervisor, supervisor backup and execute grid nodes. Each grid node has computing cost level to charge each running job. Transfer jobs from high CPU utilization grid nodes to low utilization grid nodes through supervisor scheduling, the system performance will be better. In this paper, we propose a job schedule model (JSM) base on grid environment. By this model, we can make all grid nodes be loading balance. When the load of the node is high, it sends the request to the supervisor and it selects the most suitable node to execute. Via implementing this model, we can improve the performance of grid environment efficiently.