Researching The Maintenance Phase Reliability Based Weh Logs Analysis

This paper mainly studies the reliability problems of Webapplications during the maintenance phase. This paper improves thereliability of Web applications by analyzing the Web log data. And theWeb applications obtains more detailed and comprehensive Web logdata by continuously maintain and update, supporting the reliabilityrunning of Web applications. Firstly, the paper uses ODC (OrthogonalDefect Classification) analysis method analysis and studies the Web logdefects and the cause of defect in the maintenance phase. The paperindependently classifies the different properties of the defect, analyzesthe reasons for the formation of these defects and result that will lead to,deepens the understanding of the process from the perspective of themulti-dimensional space, and locations the root of the Web applicationdefects. This paper uses the ODC orthogonal defect classificationanalysis methods analyzing the Web access log attributes according toresponse code, file type, agent type and observation time to locate theerror， and verifies the method applicability and effectiveness byexperiments. Secondly, the paper presented a regression testing datachecking and screening method by combined the widely usedregression test selection techniques and the update coverage data.This algorithm uses the record of traversal in a deep-first order effectivelyimprove the accurate of regression testing, reduces the expense ofregression testing by reducing the time and omitting the no need torerun test case, and achieves the purpose of reducing the cost ofregression testing. Finally, for Web logs couldnt record data due toexternal causes, network reason or other reasons during recordingprocess, the paper proposed a reliability of Web applications, in orderto collect as completely as possible Web log data. The paper providesreference data and effective programming methods for Web application reliability.