摘要
随着数据中心规模的扩大,提升资源利用率的问题日益凸显.如何合理部署虚拟机到数据中心的物理机上是提升数据中心资源利用率的关键问题之一.首先提出相应的资源利用率和响应时间模型,并提出一种结合变异算子的多目标粒子群优化算法(Multi-Objective Particle Swarm Optimization with Mutation Operator,MOPSOM)用于解决虚拟机部署问题.该算法以提高资源利用率和减少响应时间为优化目标,采用线性递减惯性权重寻找最优的虚拟机分配方案.仿真实验结果表明,MOPSOM能够在减小响应时间下提高资源利用率,同时在负载均衡度和能耗之间达到了更好的平衡.
With the expansion of the data center,the problem of increasing the utilization of resources is becoming increasingly prominent.How to deploy the virtual machine to the data center is one of the key issues to improve the resource utilization of the data center.Firstly,the corresponding resource utilization and response time model are proposed,and a multi-objective particle swarm optimization algorithm with mutation operator（ MOPSOM） is proposed to solve the problem of virtual machine deployment. In order to improve the efficiency of resource utilization and reduce the response time,the algorithm uses linear decreasing inertia weight to find the optimal virtual machine allocation scheme. The simulation results show that MOPSOM can improve the efficiency of resource utilization in reducing response time,and achieve a better balance between load balance and energy consumption.
作者
黄启成
陈羽中
江伟
刘耿耿
HUANG Qi-cheng;CHEN Yu-zhong;JIANG Wei;LIU Geng-geng(College of Mathematics and Computer Science, Fuzhou University, Fuzhou 350108, China;Fujian Provincial Key Laboratory of Networking Computing and Intelligent Information Processing, Fuzhou University, Fuzhou 350108, China;Fujian Collaborative Innovation Center for Big Data Applications in Governments, Fuzhou 350003, China)
出处
《小型微型计算机系统》
CSCD
北大核心
2018年第7期1554-1559,共6页
Journal of Chinese Computer Systems
基金
国家自然科学基金项目(61300102
61300103
61300104
11501114)资助
福建省自然科学基金项目(2013J01230
2014J01233
2013J01232)资助
福建省杰出青年科学基金项目(2014J06017
2015J06014)资助
福建省教育厅重点项目(JK2012003)资助
福建省科技厅高校产学合作重大项目(2014H6014)资助
福建省科技创新平台项目(2014H2005)资助
福建省科技平台建设项目(2009J1007)资助
关键词
云计算
虚拟机部署
粒子群优化
变异算子
资源利用率
响应时间
cloud computing
virtual machine placement
particle swarm optimization
mutation operator
resource utilization rate
response time
