摘要
提出云数据中心基于遗传算法的虚拟机迁移模型GA-VMM(genetic algorithm based virtual machine migration)。GA-VMM在虚拟机迁移时刻考虑的问题维度优于常见的策略,使虚拟机的分配与迁移更加合理与公平。建立了云端能量消耗与在线虚拟机迁移时间消耗数学模型,通过全局遗传算法来优化虚拟机迁移和放置策略。利用某个企业的大数据中心作为云端测试环境,对比测试GA-VMM迁移模型与已有的虚拟机迁移策略的性能。测试结果表明,GA-VMM迁移模型能够更好地减少物理主机的使用数量和虚拟机的迁移次数,SLA(service level agreement violation)违规基本处于稳定状态;GA-VMM可以降低数据中心能耗,性能优于已有的迁移策略。
This paper proposed and discussed a genetic algorithm based on virtual machine migration model in cloud data centers called GA-VMM.It took into account several important parameters to frame the objective function for the virtual machine migration which overwhelmed the old strategies and thus it made the virtual machine migration more reasonable and fair.It constructed the mathematics model of power consumption and live migration cost of virtual machine in GA-VMM.It designed a global optimization genetic algorithm for how virtual machine migrated and placed.Using a corporation cloud data centers ser-vers,this paper did some simulation experiments.The experimental results show that GA-VMM can effectively reduce the number of physical machine usages and virtual machines migrations compared with the common strategy.It had also obtained the fewer SLA violation and low energy consumption of data centers in our test.
作者
刘开南
Liu Kainan(School of Information&Intelligent Engineering,San Ya University,Sanya Hainan 572022,China)
出处
《计算机应用研究》
CSCD
北大核心
2020年第4期1115-1118,共4页
Application Research of Computers
基金
海南省自然科学基金面上资助项目(618MS082)
海南省高等学校教育教学改革研究重点资助项目(Hnjg2017ZD-19)
科技部国家重点研发计划专项资助项目(2017YFC0306400)。
关键词
低能量消耗
SLA违规
虚拟机迁移
云数据中心
遗传算法
low energy consumption
SLA violation
virtual machine migration
cloud data centers
genetic algorithm
