基于可信度策略的校园网格作业调度算法

Job scheduling algorithm based on credit strategy in campus grid

为了实现校园网上的资源共享,降低资源浪费,文中讨论了基于网格技术的校园网作业调度模型,并设计和实现了基于可信度遗传算法的作业调度方法.该方法采用资源可信度策略,同时结合遗传算法的优点,满足了网格对调度系统可扩展性和全局最优调度的需求,从而使调度系统具有了一定的自主性和智能性.实验结果表明该作业调度方法收敛速度快,全局寻优能力强,整体性能优于其他调度算法.

In order to share distributed resources in the campus network and save relevant cost, a model that uses the grid technology to schedule job is discussed, and a job scheduling method based on the CGJSA （credit genetic job scheduling algorithm） in campus grid is designed and implemented. The resource credit strategy and genetic algorithm are used as scheduling policy, which serves the need of the grid resource management scheduler well and enables this system to exhibit a certain degree of self-determination and intellectuality. From the analysis and experiment result, it is concluded that the CGJSA has the characteristics of rapid convergence, good global search capacity, and therefore is suoerior to other algorithm as a whole.