Abstract:
This paper presents a Web load test method based on log analysis. This method gets the frequent sequential patterns of Web users from sequential pattern mining. It gets the parameters of Web load from log analysis. It uses them to generate test load closed to realistic load, and uses LoadRunner to do the Web load test. In the experiment, the log generated by test is compared with realistic log, and the similarity between test load and realistic load is validated.
Key words:
log analysis,
load test,
sequential pattern mining,
PrefixSpan algorithm
摘要： 提出一种基于日志分析的Web负载测试方法。通过序列模式挖掘得到用户访问的频繁序列模式，日志分析得到负载的相关参数，并据此构造出逼近于真实的测试负载。利用性能测试工具LoadRunner对Web应用系统进行负载测试。将测试日志与真实日志进行对比，验证了测试负载与真实负载的相似性。
关键词:
日志分析,
负载测试,
序列模式挖掘,
PrefixSpan算法
CLC Number:
SHU Jing-Jun, TUN Hai-Yan, GAO Guo-Zhu, CHENG Zhi-Dui. Web Load Test Method Based on Log Analysis[J]. Computer Engineering, 2010, 36(23): 25-27.
朱靖君, 吴海燕, 高国柱, 程志锐. 一种基于日志分析的Web负载测试方法[J]. 计算机工程, 2010, 36(23): 25-27.