Min-Min任务调度算法的思路总是优先调度执行时间较短的小任务,无法得到理想的最优跨度及资源负载平衡。针对该问题,提出基于资源分级的自适应Min-Min算法。分配任务前,先参考现有资源的属性进行分级处理,再与任务在资源中的最小完成时间作乘积得到的最小任务资源组合进行调度;在任务调度过程中,引入自适应阈值,调节长任务的调度等级,从而达到优化效果。通过模拟仿真实验,表明该算法在时间跨度和负载平衡上均有较好性能。
The Min-Min task scheduling algorithm in distributed computing is simple and clear,which is always the priority scheduling perform small tasks for a short time,unable to make in respect of the optimal span and resource load balancing the ideal balance. In order to solve this problem,this paper proposed an adaptive Min-Min task scheduling algorithm based on resource classification. Firstly,it classified the existing resource according to its own properties,and then multiplied the minimum completion time by the task in the resources to get the optimal combination of tasks and resources. Secondly,it adjusted the long task priorities to achieve good performance with adaptive threshold valve. The simulation experiment results show that the algorithm has obvious improvement on the time span and load balancing.