数据中心环境下能耗性能感知的优化方法

摘要 针对云计算环境下数据中心高能耗问题,提出了一种能耗性能感知的虚拟机动态整合方法,以优化能耗性能. 首先通过局部回归分析判定主机是否过载,利用均值法检测主机是否轻载,然后根据虚拟机最小迁移时间、最大和最小CPU利用率策略选择要迁移的虚拟机加入到迁移队列,最后运用能耗性能感知的虚拟机部署优化算法对虚拟机进行优化部署. 仿真结果显示,该算法不但提高了系统性能,而且也降低了系统能耗.
In order to optimize the high energy consumption and to keep certain quality of service in the cloud computing data center, an optimization method for performance and energy aware was proposed. Firstly, local regression was used to determine whether the host is overloaded and using mean value determined whether the host is underloaded. And then, the virtual machine of the overloaded host was shifted into queue according to the minimum migration times and the maximum or minimum CPU utilization strategy. Finally, the performance and energy aware algorithm was designed to deploy the virtual machine in the queue and underloaded host. Simulations show that the proposed algorithm not only improves system performance but also reduces system energy consumption.