摘要
提出了一种测试软件老化过程的实验方案 ,编制了实现该方案的软件工具集 ,并以应用广泛的 Apache网络服务器为目标 ,研究了当负载为泊松流时服务器的老化现象 ,用回归分析和 ARIMA(自回归差分滑动平均 )模型分析了剩余内存的变化情况。结果表明 ARIMA模型可以较好地描述和预测内存的使用情况。
A new methodology for detecting software-aging phenomenon is proposed,and a suit of corresponding software tools is designed. The software-aging phenomenon in Apache Web server under Poisson-distributed workload is studied,while Regression and ARIMA models are used to validate the trend of software aging. Experimental results show that the measured data are agreement with the estimated results.
出处
《数据采集与处理》
CSCD
2004年第2期231-234,共4页
Journal of Data Acquisition and Processing
关键词
软件测试
数据采集软件
网络服务器软件
老化现象
software aging
Apache web server
regression analysis
autoregressive integrated moving average (ARIMA) model
