凸价格函数下基于堆栈执行的云计算资源调度方案

摘要

For the issue that the existing cloud resource scheduling scheme doesn't take into account the dynamic price of resources,this paper proposed a new task scheduling scheme under the convex price function.First,this scheme modeled the rent price of cloud resource as a convex function,and divided the task execution interval into multiple units.Then,it rescheduled the tasks in the low density interval to other intervals under satisfy the task deadline,which according to the task density in the sub interval,so as to make the tasks stack for execution.Finally,based on the properties of the convex price function,it allocated the maximum resource for each task with the maximum execution rate of the task,so that to get more preferential price and reduce the total cost of the leased resource.Experimental results show that the proposed scheme can effectively reduce the total cost of task execution.%针对现有云任务调度方案没有考虑到资源动态价格的问题,提出一种在凸价格函数下的资源调度方案.首先,将云资源租用价格建模为凸函数,并将任务执行区间划分为多个单位子区间;然后,根据子区间中的任务密度,在满足任务截止时间下,将低密度区间中的任务重新调度到其他区间,使任务堆栈执行;最后,根据凸价格函数的特性,以任务的最大执行速率为每个任务分配以最大资源执行,以此获得更加优惠的资源租用单价,降低用户的总成本.实验结果表明,该方案能够有效降低任务执行的总成本.