多目标优化的云计算虚拟集群动态调整方法  

Multi-Objective Optimization Method for Adaptive Adjustment of The Virtual Cluster in Cloud Computing

作　　者：张玮[1,2] 牛正浩 史慧玲[1] 孙萌[1] 唐绍炬[1] 

机构地区：[1]山东省计算中心山东省计算机网络重点实验室,山东济南250000 [2]中国互联网络信息中心互联网基础技术开放实验室,北京100190 [3]中国人民大学法学院,北京100000

出　　处：《济南大学学报（自然科学版）》2014年第5期376-381,共6页Journal of University of Jinan(Science and Technology)

基　　金：国家自然科学基金(61272433);山东省优秀中青年科学家科研奖励基金(BS2011DX033)

摘　　要：当前云计算应用中提供的虚拟集群尚不能自动的根据实际网络应用负载作动态调整,从而导致网络虚拟集群的资源利用与负载不适应,使得较多的计算资源空闲在集群中,产生能源浪费。针对上述问题,提出一种多目标优化的云计算虚拟集群动态调整方法,并给出虚拟集群动态调整框架。实验结果表明,该方法可以自动调整虚拟集群规模以适应网络应用负载,并能达到充分利用云计算资源和减少云计算能源消耗等优化目标。The virtual cluster provided in the cloud computing cannot be automatically adjusted based on the actual load in the network applications. So that makes the resource in the virtual cluster not suit to the load, and makes more idle computing resources in the cluster. For these reasons, a multi-objective optimization method of adaptive adjustment of the virtual cluster in cloud computing is proosed, and a framework of the adaptive adjustment is also provided. Experimental results show that the method can automatically adjust the virtual cluster size to fit the network application loads, and can take full advantage of the resources of cloud computing and save energy.

关 键 词：云计算 虚拟集群 多目标优化 动态调整 

分 类 号：TP393.4[自动化与计算机技术—计算机应用技术]