云环境下的基于Min-Max的节能资源调度算法的研究

ENERGY-SAVING RESOURCE SCHEDULING ALGORITHM BASED ON MIN-MAX IN CLOUD ENVIRONMENT

针对云计算环境下的高能耗问题,从系统节能的角度提出一种节能资源调度算法(energy-saving scheduling algorithm based on min-max,ESSAMM)。在Min-Max算法的基础上综合考虑了用户对于任务期望的完成时间和能量消耗两个因素,以节省任务执行过程中产生的能量消耗,并提高用户的时间QoS满意度,实现负载均衡。将任务集合中各任务按照长度从小到大排序,并根据时间QoS为该集合中长度最大和最小的任务选出符合用户期望的物理资源;根据能量估算模型,计算出这两个任务在各物理机上的执行能耗;选择最小能耗对应的物理机来执行该任务;将这两个任务在任务集合中删除,并重复上述过程,直到任务集合为空。仿真结果表明,相比于Min-Max和Min-Min资源调度算法,该算法能够有效降低系统执行任务产生的总能耗,提高用户时间服务质量,并实现调度系统负载均衡。

Aiming at the high energy consumption problem in the cloud computing environment,this paper proposes an energy-saving scheduling algorithm based on min-max(ESSAMM)from the perspective of system energy saving.On the basis of Min-Max algorithm,it takes into account two factors comprehensively:the user s completion time and energy consumption of the task expectations to save the energy consumption in the process of task execution and improve user s time QoS satisfaction,and achieves the load balancing.Each task in the collection of tasks was firstly sorted from small to large in terms of length.According to the time QoS,the maximum and minimum length in the collection of tasks were selected to meet the user s expectation of physical resources.Then,the energy consumption of the two tasks on each physical machine was calculated by the energy estimation model.We selected the physical machine corresponding to the minimum energy consumption to perform the tasks.After that,we deleted these two tasks in the collection of tasks,and repeated the above process until the collection of tasks was empty.The simulation results show that compared with the resource scheduling algorithms Min-Max and Min-Min,our algorithm can effectively reduce the total energy consumption,improve the quality of user time service,and achieve load balancing of the scheduling system.