Research Of Job Scheduling Management System Based On Grid Portal

Grid computing which organizes distributed computers as a "virtual super computer" by network provides dependable, consistent, pervasive, and inexpensive access to high-end computational capabilities. It provides an ideal platform to solve large-scale computing problems in scientific and engineering area.Job Management System is a key technology in computational grid. In grid environments, from job submission to result processing, all events about jobs are under the control of job management. Because grid environments are large-scale, heterogeneous, dynamic, distributed and autonomous, grid job management is complex and challenging. Referring current research results on job scheduling algorithms in grid environments, a job scheduling algorithm based on greedy strategy and Sufferage algorithm is proposed. The main research is described as follow:(1) Through the research of traditional job scheduling model in grid environments, considering the network delay factor, a new mathematical model of computation grid job scheduling is established. It is proposed that the resources consumption is the accumulation of resources and time. At the same time, total running time is reduced.(2) By analyzing the traditional Sufferage, Min_Max algorithm and combining mathematical model and the greedy strategy, Greedy_Sufferage algorithm is proposed. Not only detailed description and the qualitative analysis about this algorithm are given, but also the rationality and superiority for this algorithm is proved. Greedy_Sufferage algorithm is implemented and designed by Java language.(3) With the grid simulation platform, compared to the three algorithms Greedy_Sufferage, Min_Max, Sufferage, the results reflects Greedy_Sufferage done well in the average latency time and throughput rate, so the system efficiency and the resources utilization ratio is improved.(4) Based on Greedy_Sufferage algorithm, a grid work management system NJMS (Network Job Management System) to manage the resources effectively is designed. The resource node for the job is selected, assigned, scheduled and monitored. The main goal is to reinforce the job management function of OS and use the resource of system effectively. Therefore the network load of system is balanced and the performance of the whole system is improved.