基于云计算服务质量感知的虚拟机节能管理研究

Energy-efficient Virtual Machines Management Using QOS-aware in Cloud Computing

云计算是大量的虚拟化的计算机资源的服务节点,如何管理基于节能型和服务型的动态可扩展资源已成为一个重要的问题。针对这一目的,综合大量前期工作,提出了一种改进的遗传算法,并构造系统模型,通过使用CloudSim(云计算仿真软件)和CloudAnalyst(云分析软件)进行定性和定量的数据分析。同时也与传统的动态电压和频率缩放(DVFS)做了比较,通过数据验证证明出利用服务质量感知对虚拟机的节能管理在响应时间、能源消耗、虚拟机迁移数量及合并适应性方面都起到改进作用。表现为在相同功率条件下,新方法能降低用户请求的响应时间,进而提高了用户的服务质量;在相同的响应时间内,新方法又能有效的降低能量功耗。这些改进都能提高用户对服务质量的满意度,同时也为未来使用并行计算技术打下了基础。

Cloud computing is a pool of virtualized computer resources （service nodes）, how to manage the dynamically scalable resources based on the energy-efficient and Qos（response time）-aware have become a significant issue. Aiming to this purpose, an improved genetic algorithm has been proposed in this paper. Performance of the algorithm is analyzed both qualitatively and quantitatively using CloudSim and CloudAnalyst. A comparison is also made with dynamic voltage and fre- quency scaling （DVFS）. Through data validation, it is proved that the energy management of virtual machine can improve the response time, energy consumption, the number of virtual machine migration and the CF of the virtual machine. Under the same power condition, the new method can reduce the response time and improve the quality of service. In the same time, the new method ean effectively reduce the energy consumption. These improvements can improve the quality of service satisfaction, but also for the future use of parallel computing technology to lay the foundation.