基于自适应过载阈值选择的虚拟机动态整合方法

发布时间：2017-12-28 20:05

本文关键词：基于自适应过载阈值选择的虚拟机动态整合方法　出处：《计算机应用》2016年10期 　论文类型：期刊论文

【摘要】：针对云环境下动态工作负载的不确定性,提出了基于自适应过载阈值选择的虚拟机动态整合方法。为了权衡数据中心能源有效性与服务质量间的关系,将自适应过载阈值的选择问题建模为马尔可夫决策过程,计算过载阈值的最优选择策略,并根据系统能效和服务质量调整阈值。通过过载阈值检测过载物理主机,然后根据最小迁移时间原则以及最小能耗增加放置原则确定虚拟机的迁移策略,最后切换轻负载物理主机至休眠状态完成虚拟机整合。仿真实验结果表明,所提出的方法在减少虚拟机迁移次数方面效果显著,在节约数据中心能源开销与保证服务质量方面表现良好,在能源的有效性与云服务质量二者之间取得了比较理想的平衡。
[Abstract]:In view of the uncertainty of dynamic workload in the cloud environment, a virtual machine dynamic integration method based on adaptive overload threshold selection is proposed. In order to weigh the relationship between data center energy efficiency and service quality, the problem of adaptive overload threshold selection is modeled as Markov decision process, and the optimal selection strategy of overload threshold is calculated, and according to the system energy efficiency and service quality, the threshold is adjusted. Overload physical threshold is detected through overload threshold. Then the migration strategy of virtual machine is determined according to the minimum migration time principle and the minimum energy consumption increase placement principle. Finally, the virtual machine integration is completed by switching the light load physical host to the dormancy state. The simulation results show that the proposed method is effective in reducing the number of virtual machine migration, performs well in saving data center energy overhead and ensuring service quality, and achieves an ideal balance between the energy efficiency and cloud service quality between the two.
【作者单位】：江南大学物联网工程学院;物联网应用技术教育部工程研究中心(江南大学);
【基金】：江苏省科技厅产学研联合创新基金资助项目(BY2013015-23)~~
【分类号】：TP302
【正文快照】： 0引言高能耗是云数据中心资源管理面临的一大挑战,随着数据中心的规模不断扩大,高能耗问题更加突出[1]。IBM的调研报告[2]指出,数据中心物理主机的平均CPU使用率只有15%~20%,而处于空闲状态的物理主机一般消耗其峰值能耗的70%[3]。可见,云数据中心开启的空闲物理主机越多,则其

【相似文献】