面向云数据中心节能的虚拟机迁移算法研究

发布时间：2018-04-22 02:33

本文选题：云数据中心 + 资源利用率　；参考：《河南科技大学》2014年硕士论文

【摘要】：近年来，随着云计算技术的发展，越来越多的资源集中在数据中心，给数据中心能耗的高效管理带来了挑战。数据中心的高能耗不仅造成电能的浪费、系统运行的不稳定，同时也对环境造成不良影响。因此，在利用云数据中心资源的同时，也应该考虑数据中心的高能耗问题。虚拟化作为云计算的基础，在数据中心资源管理方面发挥着重要作用。如何利用虚拟化技术提高数据中心的资源利用率，降低能耗，让云数据中心更加节能成为近年来一个研究的热点。论文从能耗测量、能耗建模等方面对云数据中心的能耗问题进行了系统分析和研究，针对虚拟机能耗测量难的问题，提出了云数据中心中虚拟机能耗的监控和测量方法，建立了虚拟机系统以及虚拟机迁移时的能耗模型。常用虚拟机迁移方法使用启发式算法分配虚拟机，其求解结果易陷入局部最优解，论文在研究遗传算法的基础上，给出了一种基于遗传算法的虚拟机迁移算法(Migratingalgorithm based on Genetic Algorithm，MGA)。该算法利用遗传进化的全局搜索原理实现虚拟机到目标节点的映射，并通过将虚拟机和目标节点的资源利用率作为输入因子引入计算过程，对遗传算法中的目标函数进行改进，在满足服务级别协议的条件下，使目标节点的使用个数及迁移次数最少，从而实现数据中心的节能。在Cloudsim仿真平台中对基于遗传算法的虚拟机迁移算法进行了仿真，，并使用能耗模型对基于遗传算法的虚拟机迁移算法实现过程中的能耗进行了测量和分析。该算法与单阈值方法(ST)和双阈值方法(DT)相比，不仅提高了算法的搜索速度，并且减少了迁移次数和物理节点的使用数量，提高了数据中心的资源利用率，降低了能耗。
[Abstract]:In recent years, with the development of cloud computing technology, more and more resources are concentrated in the data center, which brings challenges to the efficient management of data center energy consumption. The high energy consumption of the data center not only causes the waste of electric energy and the unstable operation of the system, but also has a bad effect on the environment. Therefore, the high energy consumption of the data center should be taken into account when using the cloud data center resources. As the foundation of cloud computing, virtualization plays an important role in data center resource management. How to use virtualization technology to improve resource utilization, reduce energy consumption and make cloud data center more energy efficient has become a hot topic in recent years. In this paper, the energy consumption of cloud data center is systematically analyzed and studied from the aspects of energy consumption measurement and energy modeling. Aiming at the difficulty of virtual machine energy consumption measurement, the monitoring and measuring method of virtual machine energy consumption in cloud data center is put forward. The virtual machine system and the energy consumption model of virtual machine migration are established. Heuristic algorithm is used to allocate virtual machine in common virtual machine migration method, and the solution result is easy to fall into local optimal solution. Based on the research of genetic algorithm, a migration algorithm of virtual machine based on Genetic algorithm based on genetic algorithm is presented in this paper. The algorithm uses the global search principle of genetic evolution to realize the mapping of virtual machine to target node. By introducing the resource utilization ratio of virtual machine and target node into the calculation process, the objective function in genetic algorithm is improved. Under the condition of satisfying the service level agreement, the number of target nodes and the number of migrations are minimized, so that the energy saving of the data center can be realized. The virtual machine migration algorithm based on genetic algorithm is simulated in Cloudsim simulation platform, and the energy consumption of virtual machine migration algorithm based on genetic algorithm is measured and analyzed by energy consumption model. Compared with single threshold method (STT) and double threshold method (DTT), this algorithm not only improves the search speed, but also reduces the number of migrations and the number of physical nodes, improves the resource utilization of the data center and reduces the energy consumption.
【学位授予单位】：河南科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP302;TP308

【参考文献】