| Summary: | 碩士 === 中國文化大學 === 資訊管理研究所 === 97 === The functions of grid computing environment are resource (e.g., CPUs, storages, etc.) sharing, collaborative processing, reliable and secure connection, etc. However, each resource of coordinate nodes in the grid environment changes dynamically. Therefore, the load-balancing of these resources usages is an important issue. In order to have good performance in the system based on grid environment, we should have a good load-balancing mechanism of all grid nodes resources in the system.
In this paper, we propose a process scheduling analysis model based on grid environment. By this model, we can make all grid nodes be load-balancing, and enhance the integrated performance.
There are two modules in the proposed model, saying supervisor process scheduling analysis module (SPSAM) built on the supervisor grid node, and execute process scheduling analysis module (EPSAM) built on the execute grid node. SPSAM can collect the latest grid nodes information regularly, supervise the statuses of grid nodes, and select the most suitable nodes to do jobs transfer. EPSAM can transfer jobs to the nodes selected by SPSAM, and receive the executed result from the selected nodes.
Via implementing the proposed model, it can supervise the load statuses of grid nodes dynamically, and make all grid nodes be load-balancing via transferred jobs. The proposed model can reduce overall computing time, and enhance the grid environment performance efficiently.
|
|---|
