摘要
目前,传统的数据中心运行方式暴露出在资源分配、管理灵活度、资源利用率、应用服务质量等方面的问题。随着虚拟化技术的成熟,采取基于云计算架构的资源池方案,可对虚拟化资源进行统一管理、调配,达到自动化、智能化管理信息系统的目的。文章在介绍云计算、虚拟化技术的基础上,着重探究基于反向传播(Back Propagation,BP)神经网络预测模型的虚拟机池负载均衡方案,并用仿真实验验证方案的可行性。仿真实验表明,预测模型能够在误差允许的范围内准确预测物理机的负载情况,具有一定的应用价值。
Nowadays, there are some problems on resource allocation, management flexibility, resource utilization and application service quality in data center with the traditional operating mode. With the development of virtualization technology, adopting resource pool scheme based on cloud computing architecture could conduct managing and deploying virtual resources uniformly and achieve the goal of controlling information system automatically and intelligently. On the basis of introduction for cloud computing and virtualization technology, this paper studies the load balance scheme of virtual machine pool based on BP neural network predicting model and verifies its feasibility by simulation. Simulation results show that the prediction model can accurately predict the load of the physical machine in the allowed range of error, and it has a certain application value.
引文
[1]倪珍.基于云计算架构的虚拟机池管理系统[D].重庆:重庆大学,2012.
[2]SADASHIV N.Cluster,grid and cloud computing:A detailed comparison[C]//Computer Science&Education(ICCSE),20116th International Conference on.Singapore:2011:477-482.
[3]肖云鹏,刘宴兵.云计算关键技术与应用展望[J].数字通信,2010(6):29-30.
[4]岳冬利,刘海涛,孙傲冰.Iaa S公有云平台调度模型研究[J].计算机工程与设计,2011,32(6):1889-1892.YUE Dong-li,LIU Hai-tao,SUN Ao-bing.Research on Iaa Spublic cloud computing platform scheduling model[J].Computer Engineering and Design,2011,32(6):1889-1892.
[5]樊邦奎,丁冠军,兰海滨,等.面向智能电网应用的云计算架构研究[J].电力信息与通信技术,2014,12(1):1-6.FAN Bang-kui,DING Guan-jun,LAN Hai-bin,et al.Research on cloud computing architecture for smart grid applications[J].Electric Power Information and Communication Technology,2014,12(1):1-6.
[6]ADABALA S,CHADHA V,CHAWLA P,et al.From virtualized resources to virtual computing grids:the In-VIGO system[J].Future Generation Computer System,2005,21(6):896-909.
[7]申向阳,丁翔.服务器虚拟化应用的研究与实践[J].石油化工自动化,2012,48(3):40-42.SHEN Xiang-yang,DING Xiang.Research and practice of server virtualization application[J].Automation in Petro-Chemical Industry,2012,48(3):40-42.
[8]张华峰,韩旭杉,王刚.基于开源虚拟机管理系统的资源整合研究与应用[J].电力信息与通信技术,2015,13(2):96-100.ZHANG Hua-feng,HAN Xu-shan,WANG Gang.Research and application of resources consolidation based on open source virtual machine management system[J].Electric Power Information and Communication Technology,2015,13(2):96-100.
[9]张笑燕,王敏讷,杜晓峰.云计算虚拟机部署方案的研究[J].通信学报,2015,36(3):084(1-8).ZHANG Xiao-yan,WANG Min-ne,DU Xiao-feng.Research on the method of virtual machine deployment in cloud computing[J].Journal of Communications,2015,36(3):084(1-8).
[10]黄辉,刘秋让.基于VMware v Sphere的图书馆服务器虚拟化研究与应用[J].当代图书馆,2011(4):41-43.HUANG Hui,LIU Qiu-rang.The research and application of server virtualization in Library based on VMware v Sphere[J].Contemporary Library,2011(4):41-43.
[11]侯利春.云环境下虚拟机动态迁移触发策略分析[J].电子信息与软件工程,2014(17):102-103.
[12]刘鑫.云计算环境下电力网络负载均衡调度方法研究[J].电网与清洁能源,2015,31(9):63-66.LIU Xin.Study on power network load balance scheduling method in cloud computing environment[J].Power System and Clean Energy,2015,31(9):63-66.
[13]余向前,路民辉,任琳杰,等.基于改进型Elman神经网络的短期电力负荷预测[J].电力信息与通信技术,2014,12(2):39-43.YU Xiang-qian,LU Min-hui,REN Lin-jie,et al.Short-term load forecast based on improved Elman neural network[J].Electric Power Information and Communication Technology,2014,12(2):39-43.
[14]陈荣征,李广松.基于遗传神经网络的主机负载预测方法研究[J].计算机时代,2009(10):12-17.CHEN Rong-zheng,LI Guang-song.Research on host prediction based on genetic neural network[J].Computer Era,2009(10):12-17.