• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

计算机工程与科学 ›› 2020, Vol. 42 ›› Issue (12): 2109-2116.

• 高性能计算 • 上一篇    下一篇

申威架构下的虚拟机访存特征提取方法

沙赛1,王超2,杜翰霖1,罗英伟1,汪小林1,王振林3   

  1. (1.北京大学信息科学技术学院,北京100871;2.江南计算技术研究所,江苏 无锡 214083;

    3.密歇根理工大学计算机系,霍顿 密歇根州 49246)

  • 收稿日期:2020-04-06 修回日期:2020-06-28 接受日期:2020-12-25 出版日期:2020-12-25 发布日期:2021-01-04
  • 基金资助:
    国家重点研发计划(2018YFB1003604);国家自然科学基金(61472008,61672053,U1611461)

A virtual machine memory access feature extraction method under Sunway architecture

SHA Sai1,WANG Chao2,DU Han-lin1,LUO Ying-wei1,WANG Xiao-lin1,WANG Zhen-lin3   

  1. (1.School of Electronics Engineering and Computer Science,Peking University,Beijing 100871,China;

    2.Jiangnan Institute of Computing Technology,Wuxi 214083,China;

    3.Department of Computer Science,Michigan Technological University,Michigan 49246,USA)

  • Received:2020-04-06 Revised:2020-06-28 Accepted:2020-12-25 Online:2020-12-25 Published:2021-01-04

摘要: 虚拟化技术是云服务的重要支柱之一,虚拟化充分扩展了物理资源的灵活性,提升了物理资源的利用率。随着国家信息化水平的发展,云服务器核心技术自主可控、安全高效的要求不断提高。近年来,作为国产服务器的典型代表,申威架构服务器的功能不断完善。提出了申威架构上的虚拟机访存特征提取方法,充分利用了申威架构独特优势,实时测算虚拟机的内存缺失率曲线,并最终计算工作集大小,同时利用热页集机制大幅度减少页面追踪的性能开销。实验结果表明,该方法可以准确计算虚拟机工作集大小,平均误差低于3%,平均性能开销不高于8.3%。本工作为申威虚拟机内存动态分配提供条件,最终目标是提高申威云服务器整体性能和内存利用率。


关键词: 申威, 虚拟化, 内存, 工作集, 内存缺失率曲线, 热页集

Abstract: Virtualization technology is one of the important pillars of cloud services. Virtualization fully extends the flexibility of physical resources and improves the utilization of physical resources. With the development of national informatization level, the core technology of cloud servers is more and more independent, controllable, safe and efficient. In recent years, as a typical representative of domestic servers, Sunway has been fully developed. In order to further improve the performance of virtual machines in Sunway architecture, this paper proposes a method of extracting the memory access features of virtual machines in Sunway architecture. By combining the advantages of Sunway architecture, this paper calculates the memory page miss ratio curve of virtual machines online based on the least recently used stack method, and reduces the performance cost of feature by using the hot set mechanism. Experiments show that the method can accurately calculate the working set size of virtual machine, the average error is less than 3%, and the average performance cost is no more than 8.3%. This work provides conditions for dynamic memory allocation of Sunway virtual machines, and finally improves the overall performance and memory utilization of Sunway cloud servers.





Key words: Sunway architecture, virtualization, memory, working set size, page miss ratio curve, hot set