计算机应用 ›› 2017, Vol. 37 ›› Issue (9): 2547-2550.DOI: 10.11772/j.issn.1001-9081.2017.09.2547

• 先进计算 • 上一篇    下一篇

基于动态调整阈值的虚拟机迁移算法

赵春, 闫连山, 崔允贺, 邢焕来, 冯斌   

  1. 西南交通大学 信息科学与技术学院, 成都 611756
  • 收稿日期:2017-02-26 修回日期:2017-04-26 出版日期:2017-09-10 发布日期:2017-09-13
  • 通讯作者: 闫连山,lsyan@home.swjtu.edu.cn
  • 作者简介:赵春(1992-),女,四川成都人,硕士研究生,主要研究方向:数据中心节能调度;闫连山(1971-),男,山东招远人,教授,博士,主要研究方向:光通信、物联网、软件定义网络;崔允贺(1987-),男,山东济宁人,博士研究生,主要研究方向:软件定义网络、网络安全;邢焕来(1983-),男,河北唐山人,副教授,博士,主要研究方向:进化计算、网格编码;冯斌(1991-),男,四川成都人,硕士研究生,主要研究方向:数据中心带宽分配。
  • 基金资助:
    国家自然科学基金资助项目(61401374);中国铁路总公司重大项目(2016X008-D)。

Dynamic adjusting threshold algorithm for virtual machine migration

ZHAO Chun, YAN Lianshan, CUI Yunhe, XING Huanlai, FENG Bin   

  1. School of Information Science and Technology, Southwest Jiaotong University, Chengdu Sichuan 611756, China
  • Received:2017-02-26 Revised:2017-04-26 Online:2017-09-10 Published:2017-09-13
  • Supported by:
    This work is partially supported by the National Natural Science Foundation of China (61401374), the Major Projects of China Railway (2016X008-D).

摘要: 针对当前数据中心服务器能耗优化和虚拟机迁移时机合理性问题,提出一种基于动态调整阈值(DAT)的虚拟机迁移算法。该算法首先通过统计分析物理机历史负载数据动态地调整虚拟机迁移的阈值门限,然后通过延时触发和预测物理机的负载趋势确定虚拟机迁移时机。最后将该算法应用到实验室搭建的数据中心平台上进行实验验证,结果表明基于DAT的虚拟机迁移算法比静态阈值法关闭的物理机数量更多,云数据中心能耗更低。基于DAT的虚拟机迁移算法能根据物理机的负载变化动态迁移虚拟机,达到提高物理机资源利用率、降低数据中心能耗、提高虚拟机迁移效率的目的。

关键词: 数据中心, 能耗, 虚拟机迁移, 预测, 资源利用率

Abstract: Aiming at the optimization of servers' energy consumption in data center and the reasonable migration time of Virtual Machine (VM), a VM migration algorithm based on Dynamic Adjusting Threshold (DAT) was proposed. Firstly, the migration threshold was dynamically adjusted by analyzing the historical load data acquired from Physical Machine (PM), then the time for migrating VMs was decided by the delay trigger mechanism and the PM load trend prediction. The VM migration algorithm based on DAT was tested on datacenter platform in the laboratory. Experimental results indicate that compared with the static threshold method, the number of the shut down PMs of the proposed algorithm is larger, and the energy consumption of the data center is lower. The VM migration algorithm based on DAT can dynamically migrate VMs according to the variation of PM load, thus improving the utilization of resources and the efficiency of VM migration, reducing the energy consumption of the data center.

Key words: data center, energy consumption, Virtual Machine (VM) migration, prediction, resource utilization rate

中图分类号: