基于互斥条件的云数据中心虚拟机整合策略

发布时间：2018-06-03 15:47

本文选题：互斥条件 + 虚拟机整合　；参考：《华东理工大学学报(自然科学版)》2017年01期

【摘要】：在云数据中心,虚拟机整合(VMC)是绿色计算和最小化集群功耗问题的关键技术。大多数研究采用基于资源使用率来寻求最合理的虚拟机(VM)整合方式,但在整合过程中并未考虑物理服务器(PM)的可用性及同位VM的稳定性问题。本文提出了基于互斥条件限制的VMC策略,该策略同时考虑PM的可用性和VM之间的互斥性两方面因素;给出了集群服务器统一资源预留计算方法,用于保证PM可用性的问题,并给出了基于VM历史运行数据的相似度判定方法,用于解决互斥条件的判定问题。将该策略应用于运行在OpenStack平台中的VM数据,实验结果表明:该策略可以有效地保证PM的可用性,避免同类型VM被整合在一起,减少同位VM的性能损失和确保服务质量。
[Abstract]:In the cloud data center, virtual machine integration (VMC) is a key technology for green computing and minimization of cluster power problems. Most studies use resource utilization to seek the most reasonable VM integration method, but in the process of integration, the availability of physical server (PM) and the stability of the same bit VM are not considered. Based on the VMC strategy of mutual exclusion conditions, this strategy considers both the availability of PM and the mutual exclusion between VM, and gives the unified resource reservation calculation method of cluster server, which is used to guarantee the availability of PM, and gives a similarity determination method based on the data of VM history running, which can be used to solve the decision of mutual exclusion conditions. This strategy is applied to the VM data running on the OpenStack platform. The experimental results show that the strategy can effectively guarantee the availability of PM, avoid the integration of the same type of VM, reduce the performance loss of the same bit VM and ensure the quality of service.
【作者单位】：华东理工大学计算机科学与工程系;上海市计算机软件测评重点实验室;
【基金】：国家自然科学基金(61173048,61300041,61472139) 高等学校博士学科点专项科研基金博导类资助课题(20130074110015) 中央高校基本科研业务费专项基金(WH1314038,WH1514331)
【分类号】：TP308

【相似文献】