Audit Trail Analysis for Traffic Intensive Web Application

Web-enabled business processes designed for travel industry and government agencies are considered as traffic intensive applications since number of users can fluctuate dramatically within a short time. Applications under this category are likely to encounter flash crowd situation in which the server can no longer handle overwhelming service requests. To alleviate this problem, business process engineers usually analyze audit trail data recorded from the application server and reengineer their processes to withstand such situations. However, audit trail analysis can only reveal some of the performance indicators which can only be observed from the internal perspective of the application server. In this research, we propose a novel approach for identifying key performance indicators of traffic intensive web applications by integrating analysis results from audit trail data of an application server with analysis results from a Web server log analyzer and stress testing tool. Web server log analyzers are mainly used to analyze the transactions between client computers and the Web server whereas stress testing tools are usually used to estimate the workload limit of a Web server. The analysis result from these programs provides an internal as well as an external view of the application performance. Such integration allows process engineers to exactly pinpoint the potential bottlenecks in the application.