Research on performance loss of virtualization
-
摘要: 虚拟化是云计算领域的一项关键技术,它将信息资源重新进行了定义和划分,解决了信息资源在调度、分配、隔离等各方面的问题,很大程度提高了信息资源的利用率。不同的虚拟化技术由于实现方式的不同,各自在性能损耗方面会有些差异。从架构和功能两方面详细介绍基于内核的虚拟机(KVM,Kernel-based Virtual Machine)和VMware两种虚拟化技术。通过几款流行的性能测试工具分别对物理机、KVM及VMware进行了性能测试,测试项包括CPU、内存、磁盘和网络。通过测试结果对比,发现两种虚拟机在硬盘驱动器(HDD,Hard Disk Driver)随机读时,会有10%的性能损耗;VMware虚拟机在作为传输控制协议(TCP,Transmission Control Protocol)客户端时,会有10%的带宽损耗;KVM在作为用户数据报协议(UDP,User Datagram Protocol)服务器时会有60%的带宽损耗;在作为UDP客户端时,会有20%的带宽损耗。基于测试结果,在实际应用中,可以根据侧重,选择合适的虚拟化技术。Abstract: Virtualization is a key technology in the field of cloud computing. It redefines and divides information resources, solves the problems of information resources in scheduling, allocation, isolation and other aspects, and greatly improves the utilization of information resources. Different virtualization technologies have some differences in performance loss due to different implementation methods.This article introduced two virtualization technologiesKernel-basedVirtual Machine (KVM) and VMware in detail from two aspects of architecture and function, used several popular performance testing tools to test the performance of physical machine, KVM and VMware, including CPU, memory, disk and network.Through the comparison of test results, it is found that two kinds of virtual machine will have 10% performance loss when the HDD(Hard Disk Driver) is read randomly; VMware virtual machine will have 10% bandwidth loss when it is used as TCP (Transmission Control Protocol) client; KVM virtual machine will have 60% bandwidth loss when it is used as UDP (User Datagram Protocol) server; and 20% bandwidth loss when it is used as UDP client. Based on the test results, we can choose the appropriate virtualization technology according to different emphasisin practical application.
-
Keywords:
- virtualization /
- performance test /
- VMware /
- KVM(Kernel-basedVirtual Machine)
-
-
[1] 孙杰,山金孝,张亮,等. 企业私有云建设指南[M]. 北京:机械工业出版社, 2019:18-19. [2] 丁勇. 云计算和虚拟化技术研究[J]. 中国新通信, 2019, 21(14):70. [3] MELL P, GRANCE T. The NIST definition of cloud computing[R]. Gaithersburg:NIST, 2012.
[4] Machine. Jin Hai, Liao Xiaofei. Virtualization Technology for Computing System[J]. China Basie Scienee, 2008, 10(6):12.
[5] 李雁明,刘相坤,李琪,等. 对客票系统运用存储虚拟化技术的研究[J]. 铁路计算机应用, 2016, 25(4):28-31. [6] 张亚龙. 基于VMware vSphere虚拟架构的服务器虚拟化系统部署-以中国-东盟多语种全媒体融合云平台为例[J]. 视听, 2019(8):216-217. [7] DANIEL P. BOVET, MARCO CESATI. 深入理解LINUX内核[M]. 陈莉君,张琼声,张宏伟,译. 北京:中国电力出版社, 2007. [8] 刘婷婷. 基于X86PC服务器虚拟化解决方案[J]. 铁路计算机应用, 2012, 21(2):42-43. [9] 王达. 深入理解计算机网络[M]. 北京:机械工业出版社, 2015. [10] 吕梅桂. 基于KVM的虚拟机网络性能优化研究[J]. 软件导刊, 2017, 16(10):180-181 , 184.
-
期刊类型引用(3)
1. 曹海鹏,唐伟忠,宋哲超,刘海宁,李思维. 车务段安全生产指挥中心管理系统设计. 铁路计算机应用. 2024(06): 46-56 . 本站查看
2. 林楷. 关于铁路制动铁鞋北斗定位防盗系统的研究. 数码世界. 2020(01): 277 . 百度学术
3. 张亮,赵明. 车务段安全生产指挥中心现状及作用提升分析. 中国铁路. 2020(07): 73-77 . 百度学术
其他类型引用(1)
计量
- 文章访问数: 606
- HTML全文浏览量: 29
- PDF下载量: 70
- 被引次数: 4