云计算虚拟资源优化分配的研究

发布时间：2018-04-14 17:26

  本文选题：云计算 + 虚拟资源分配　； 参考：《福建师范大学》2012年硕士论文

【摘要】：云计算被认为是下一代的IT服务模式,受到学术界和工业界的巨大关注。使用云计算,只需通过计算和存储能力都十分有限的终端设备,就可获得近乎无限的计算和存储能力。社会对云计算需求的不断扩大需要构建规模巨大的数据中心,而维护其运行需要大量的能量。如何高能效地运行数据中心是一个急待解决的问题。 本文首先对云计算的一些基本概念,分类,特点和研究热点等做了简单的描述,并介绍了国内外关于高能效虚拟资源分配的研究,以及目前仍存在的不足之处。其次,在服务器性能指标约束的前提下,为了有效降低服务器集群的能耗。本文将云计算中的虚拟资源分配问题模型化为一个路径构建的问题,利用改进的精华策略蚂蚁系统(elitist strategy for ant system, EAS)来进行资源分配方案的优化,仿真结果表明,在提高服务器利用率的同时,新策略有效地降低了数据中心的能耗。为了更大限度地降低数据中心的能耗,本文使用动态调压调频技术DVFS (Dynamic Voltage and Frequency Scaling),进一步将云计算中的虚拟资源分配问题建模为 个多目标优化模型,并使用优秀的多目标进化算法non-dominated sorting genetic algorithm Ⅱ(NSGA-Ⅱ)求解该多目标优化模型,仿真结果表明,针对不同特征的虚拟主机和服务器的需求,策略能够在合理的时间内生成调度方案,从而有效地降低运行数据中心所需能耗。在本文的最后部分,针对原始NSGA-Ⅱ算法在进化过程中难以通过有效的变异过程来提高对局部解空间的搜索能力,导致在求解多目标优化问题时出现的早熟现象,通过改进NSGA-Ⅱ算法的变异过程,提高算法对局部解空间的搜索能力,保持了种群的多样性,避免了早熟现象,同时改进了多目标优化模型,引入开关服务器代价做为优化目标,仿真结果表明策略能在保证SLAs (Service Level Agreements)的前提下,更加高能效地运行云计算数据中心。
[Abstract]:Cloud computing is regarded as the next generation of IT service model, which has attracted great attention from academia and industry.With cloud computing, almost unlimited computing and storage capabilities can be obtained by terminal devices with very limited computing and storage capabilities.The increasing demand for cloud computing requires the construction of a large data center, and the maintenance of its operation requires a great deal of energy.How to run the data center efficiently is an urgent problem to be solved.In this paper, the basic concepts, classification, characteristics and research focus of cloud computing are briefly described, and the domestic and foreign research on the allocation of energy efficient virtual resources, as well as the existing deficiencies are introduced.Secondly, in order to reduce the energy consumption of the server cluster effectively, under the constraint of the server performance index.In this paper, the virtual resource allocation problem in cloud computing is modeled as a path construction problem, and the improved essence strategy ant system, named as strategy for ant system, is used to optimize the resource allocation scheme. The simulation results show that,At the same time, the new strategy can effectively reduce the energy consumption of the data center.In order to reduce the energy consumption of the data center to a greater extent, this paper uses the dynamic Voltage and Frequency scaling technology DVFS to model the virtual resource allocation problem in cloud computing as follows:Non-dominated sorting genetic algorithm 鈪，

本文编号：1750329