基于二次指数平滑预测的虚拟机调度方法研究*

针对数据中心的高能耗问题，提出了一种基于负载感知和预测的虚拟机调度方法，采用二次指数平滑法预测物理主机资源负载情况，利用MMT和MM相结合的策略选择待迁虚拟机，使用资源最佳适配策略（BRF）选择目标物理主机。该调度方法的预测模型能提高迁移触发准确率，随着调度轮数的增加，对资源需求互补的虚拟机会被整合到相同物理主机上，从而减少迁移次数；最后，通过CloudSim仿真平台与FT_MMT、CDLC、AR_MMT调度策略进行了对比，结果表明该调度方法在能耗节约、迁移次数方面均有提升。

Research on virtual machine scheduling method based on double exponential smoothing prediction

Aiming at the problem of high energy consumption in data centers, this paper proposed a virtual machine scheduling method based on load prediction, which used double exponential smoothing approach to predict the load of the host, hybrid strategy of MMT and MM to choose the virtual machine which should be moved, and the best resources fit strategy (BRF) to select the physical host. The prediction model of the proposed scheduling method can improve the accuracy of triggered migration. With the increase of the number of scheduling rounds, virtual machines, which have complementary demand for resources , will be integrated into the same physical host and reduce the times of migration thereby. Finally, this paper compared the method with FT_MMT, CDLC and AR_MMT strategy on the CloudSim simulation platform. The results show that the scheduling method has improvement in energy saving and migration times reduction.