基于性能测试的信息系统监测与调优方法研究

System monitoring and optimization method based on performance testing

1. Institute of Computing Technologies, China Academy of Railway Sciences Corporation Limited, Beijing 100081, China;
2. Beijing Jingwei Information Technologies Co. Ltd., Beijing 100081, China

摘要: 为了更准确地发现并解决信息系统性能瓶颈,研究了在性能测试过程中信息系统检测与调优的技术。将系统性能测试分为CPU、内存、I/O和网络4个部分,进行原理分析和监测方法的探索,研究系统产生瓶颈的原因,并通过相应的调优方法使系统的处理能力和响应时间达到最优,实现系统在大并发压力下的平稳运行,提升用户体验度。
- 性能测试 /
- 系统测试 /
- 系统调优
Abstract: In order to find and solve the performance bottleneck of information system more accurately,this article studied the technology of information system detection and optimization in the process of performance testing.The system performance test was divided into four parts: CPU, memory, I/O and network. The article explored the principle analysis and monitoring methods, and studied the reasons for the bottleneck of the system. The processing capacity and response time of the system are optimized by corresponding optimization methods, which makes the system run smoothly under the pressure of large concurrency and improves the user experience.
- performance testing /
- system monitoring /
- performance optimization

[1]	李天翼. 12306互联网售票系统测试的实现[J]. 铁路计算机应用, 2016, 25(10):27-30.
[2]	陈绍英,周志龙,金成姬. 大型IT系统性能测试入门经典[M]. 北京:电子工业出版社, 2016.
[3]	陈能技,郭柏雅. 性能测试诊断分析与优化[M]. 北京:电子工业出版社, 2012.
[4]	修佳鹏,杨正求,刘雪梅. 软件性能测试及工具应用[M]. 北京:清华大学出版社, 2014.
[5]	于澎,刘文韬,朱建生,等. goTicket性能测试平台的开发与应用[J]. 铁路计算机应用, 2015, 24(6):1-4.
[6]	William Stallings. 计算机组成与体系结构:性能设计[M]. 北京:机械工业出版社, 2011.
[7]	Dong L, Jing X, Chunhui Y. Study of Performance Testing of Information System Based on Domestic CPU and OS[C]//Third International Conference on Trustworthy Systems & Their Applications. IEEE, 2016.
[8]	Andrew S.Tanenbaum, David J.Wetherall. 计算机网络[M]. 北京:清华大学出版社, 2012.
[9]	Evi Nemeth, Garth Snyder. UNIX/Linux系统管理技术手册[M]. 北京:人民邮电出版社, 2012.
[10]	Vincenzo Ferme, Cesare Pautasso. A declarative approach for performance tests execution in continuous software development environments[C]//ACM/SPEC International Conference on Performance Engineering, 2018, 3:261-272