云计算中基于能耗优化的虚拟机多目标放置算法

Virtual Machine Multi-objective Placement Algorithm Based on Optimization of Energy Consumption in the Cloud Computing

如何将虚拟机放置部署到合适的物理结点上是云计算提高系统性能,兼顾节能减排的关键技术之一.首先研究和建立虚拟机放置模型,为能耗优化提供基础,综合考虑应用系统性能要求、资源开销、能源消耗、虚拟机迁移等因素,将资源控制和能耗控制结合起来,形成虚拟机放置能耗优化模型,在此基础上提出基于量子多目标进化的虚拟机放置算法(EAPC),并采用二级递阶分解结构优化求解过程,实现目标的全局最优化.对比实验结果表明算法能有效地减少能源消耗,同时能较好的控制资源开销,达到节能减排,绿色计算的目的,具有较强的理论和现实意义.

In cloud computing, how to place the virtual machine deployment on physical node is one of the key technology to improve system performance and energy saving. This research provided a basis for energy consumption optimization by creating a virtual machine placing model; and by comprehensive consideration of the application performance requirements, resource consumption, energy consumption, physical node number and other factors, it is to combine the resource control and power control to form a virtual machine placing model with energy optimization; then based on this model, by using EAPC and employing two hierarchical decomposition structure to optimize the solution process and achieve an optimized goal. The experiment results show that the algorithm can effectively reduce energy consumption and can control the spending of resource to achieve the purpose of saving energy and green computing. Thus the research is of important theoretical and practical significance.