一种基于qemu的动态迁移模型
发布时间：2019-04-20 17:10
【摘要】：在线迁移已经成为数据中心的一个核心管理工具,广泛用于负载平衡、服务器整合和系统维护等方面。精确地预测在线迁移性能是制定有效迁移决策的前提。在广泛用于开源云计算的qemu-kvm虚拟化平台中,迁移策略与传统的预拷贝策略存在差异,导致已有的迁移模型无法有效地应用于该平台。为此,提出一种基于qemu-kvm平台的迁移策略的建模方法,基于模型抽取影响在线迁移性能的关键因素,分析它们与迁移性能之间的数学关系,最后针对这些关键参数建立相应的测试环境,以此测试评估模型的正确性与精确性。测试结果表明模型预测迁移时间和迁移数据总量的精确度在95%以上。
[Abstract]:Online migration has become a core management tool for data centers, widely used in load balancing, server consolidation and system maintenance. Accurate prediction of on-line migration performance is a prerequisite for making effective migration decisions. In the qemu-kvm virtualization platform widely used in open source cloud computing, the migration strategy is different from the traditional pre-copy strategy, which leads to the existing migration model can not be effectively applied to the platform. Therefore, a modeling method of migration strategy based on qemu-kvm platform is proposed. The key factors affecting on-line migration performance are extracted based on the model, and the mathematical relationship between them and migration performance is analyzed. Finally, according to these key parameters, the corresponding test environment is established to test and evaluate the correctness and accuracy of the model. The test results show that the accuracy of the model to predict migration time and total migration data is more than 95%.
【作者单位】： 重庆理工大学计算机科学与工程学院;西安交通大学计算机系;
【基金】：国家自然科学基金(61173040)资助
【分类号】：TP302
[Abstract]:Online migration has become a core management tool for data centers, widely used in load balancing, server consolidation and system maintenance. Accurate prediction of on-line migration performance is a prerequisite for making effective migration decisions. In the qemu-kvm virtualization platform widely used in open source cloud computing, the migration strategy is different from the traditional pre-copy strategy, which leads to the existing migration model can not be effectively applied to the platform. Therefore, a modeling method of migration strategy based on qemu-kvm platform is proposed. The key factors affecting on-line migration performance are extracted based on the model, and the mathematical relationship between them and migration performance is analyzed. Finally, according to these key parameters, the corresponding test environment is established to test and evaluate the correctness and accuracy of the model. The test results show that the accuracy of the model to predict migration time and total migration data is more than 95%.
【作者单位】： 重庆理工大学计算机科学与工程学院;西安交通大学计算机系;
【基金】：国家自然科学基金(61173040)资助
【分类号】：TP302
【参考文献】
相关期刊论文 前2条
1 张翔;霍志刚;马捷;孟丹;;虚拟机快速全系统在线迁移[J];计算机研究与发展;2012年03期
2 陈阳;怀进鹏;胡春明;;基于内存混合复制方式的虚拟机在线迁移机制[J];计算机学报;2011年12期
【共引文献】
相关期刊论文 前10条
1 李鸿健;代宇;刘锐;蒋溢;;云数据中心高能效的虚拟机迁移整合算法研究[J];电信科学;2015年01期
2 闫永明;张斌;郭军;莫玉岩;;基于模糊层次分析的虚拟机热度综合评估方法[J];东北大学学报(自然科学版);2015年02期
3 肖鹏;刘洞波;屈喜龙;;云计算中基于能耗比例模型的虚拟机调度算法[J];电子学报;2015年02期
4 常德成;徐高潮;;虚拟机动态迁移方法[J];计算机应用研究;2013年04期
5 张伟;张晓霞;王汝传;;一种基于脏页面延迟拷贝的虚拟机动态内存迁移方法[J];计算机科学;2013年05期
6 姚华超;王振宇;;基于KVM-QEMU与Libvirt的虚拟化资源池构建[J];计算机与现代化;2013年07期
7 魏坤;许晖;杨e，
本文编号：2461794
本文链接：https://www.wllwen.com/kejilunwen/jisuanjikexuelunwen/2461794.html
