|
|||||
|
|
| 桌面虚拟化环境下的交互式性能优化 | |
| 引用本文： | 汪小林,周凯,张彬彬,杨亮,罗英伟,李晓明. 桌面虚拟化环境下的交互式性能优化[J]. 计算机科学与探索, 2012, 6(4): 289-300 |
| 作者姓名： | 汪小林 周凯 张彬彬 杨亮 罗英伟 李晓明 |
| 作者单位： | 1. 北京大学信息科学技术学院,北京,100871 2. 北京大学深圳研究生院深圳市云计算关键技术与应用重点实验室,广东深圳,518055 3. 云南大学信息学院计算机科学与工程系,昆明,650091 4. 北京大学信息科学技术学院,北京100871;北京大学深圳研究生院深圳市云计算关键技术与应用重点实验室,广东深圳518055 |
| 基金项目： | 国家自然科学基金No.61170055;国家高技术研究发展计划(863)No.2012AA010905;高等学校博士学科点专项科研基金博导类课题No.20110001110101;教育部-英特尔信息技术专项科研基金No.MOE-INTEL-10-06;深圳市生物、互联网、新能源、新材料产业发展专项资金项目No.JC201104210107A~~ |
| 摘 要： | 硬件辅助虚拟化的提出,极大地提高了全虚拟化的性能。但是全虚拟化下类似网络教室等桌面虚拟化应用,在交互式性能方面还存在较大局限。交互式性能主要受I/O性能的影响:一方面I/O设备属于慢速设备;另一方面全虚拟化中采用模拟方式来共享设备。对全虚拟化的交互式性能进行了改进,充分利用多核处理器的物理特征来部署配置虚拟机,根据用户行为动态调整虚拟机的使用资源,依据虚拟化特征进行I/O调度的优化。最后,通过评测实验证明了改进方案的有效性。 |
| 关 键 词： | 多核 桌面虚拟化 交互式性能 I/O特征 调度协同 |
Interactive Performance Optimization for Desktop Virtualization Environment |
|
| WANG Xiao-lin , ZHOU Kai , ZHANG Bin-bin , YANG Liang , LUO Ying-wei , LI Xiao-ming. Interactive Performance Optimization for Desktop Virtualization Environment[J]. Journal of Frontier of Computer Science and Technology, 2012, 6(4): 289-300 | |
| Authors: | WANG Xiao-lin ZHOU Kai ZHANG Bin-bin YANG Liang LUO Ying-wei LI Xiao-ming |
| Affiliation: | 1.School of Electronic Engineering and Computer Science,Peking University,Beijing 100871,China 2.The Shenzhen Key Lab for Cloud Computing Technology and Applications,Shenzhen Graduate School,Peking University,Shenzhen,Guangdong 518055,China 3.Department of Computer Science and Engineering,School of Information Science and Engineering,Yunnan University,Kunming 650091,China |
| Abstract: | Full virtualization is vastly applied in desktop virtualization.Although hardware-assisted virtualization greatly improves the performance,the interactive performance is still a bottleneck for full virtualization.Interactive performance is mainly influenced by I/O devices.In one hand,I/O devices are slow device.In another hand,they are often shared by multiple virtual machines through simulation.This paper improves the interactive performance of full virtualization.Firstly,it investigates virtual machine deployment to ensure stable performance of the whole system and individual virtual machine.Secondly,it dynamically adjusts the resource among virtual machines based on the interactive behavior of individual machine.Thirdly,it optimizes the I/O request scheduler in term of the virtualization implementation.Finally,the experimental results prove the validity of the improved method. |
| Keywords: | multi-core desktop virtualization interactive performance I/O feature coordinated scheduling |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |