摘要
IaaS云计算平台中主要通过实时迁移多台虚拟机来实现资源的动态调度、管理与优化。虽然Pre-copy和Post-copy是单虚拟机实时迁移的两种主流算法,且各有优缺点,但现有的多虚拟机实时迁移系统只是单一地使用其中一种迁移算法,无法根据各虚拟机的不同负载情况灵活选择最有效的迁移算法,降低了整体迁移效率。提出一种自适应的实时迁移算法选择框架,利用模糊聚类方法对待迁移的多虚拟机进行分类,按类别选择最适合的迁移算法。实验结果表明,所提出的迁移算法选择框架能够在多虚拟机实时迁移中发挥两个迁移算法的各自优势,有效提高整体的实时迁移性能。
In IaaS cloud computing platform, live migration of multiple virtual machines plays a dominant role in the dy- namic scheduling, optimization and management of IT resources. Although Pre-copy and Post-copy are the prevalent live migration algorithms for the single virtual machine, which both have the pros and cons, only one of them is monotonous- ly adopted in the context of the gang of live migration. This scheme cannot choose the best migration algorithm for each virtual machine according to its overload, eventually degrading the whole migration performance. This paper proposed an adaptive live migration algorithm choosing framework, which uses fuzzy clustering method to classify the virtual ma- chines to be migrated and migrates them with the chosen optimum migration algorithm. Experiment results show that the proposed framework can exert each advantage of the two basic migration algorithms and improve the whole live mi- gration performance.
作者
崔勇
林予松
刘炜
高山
王宗敏
CUI Yong LIN Yu-song LIU Wei GAO Shan WANG Zong-min(Institute of Information Engineering, Zhengzhou University, Zhengzhou 450001, China Henan Provincial Key Lab on Information Networking, Zhengzhou University,Zhengzhou 450052, China)
出处
《计算机科学》
CSCD
北大核心
2016年第8期60-65,共6页
Computer Science
基金
教育部博士点专项科研基金(20114101110007)
河南省科研重点项目(13A520562)
河南省创新人才项目(2011HASTIT003)
河南省高等学校重点科研项目(15A520028)
河南省基础与前沿技术研究项目(152300410047)资助