Lbd4ws: Log-based Diagnosis for Web Service

With the development of web service application, it is highly desirable to quickly and exactly localize the service faults and explain the faulty reason. In this paper, we propose a log-based diagnosis method (LBD4WS) which combines MBD method with the statistical techniques based on historical data to construct the diagnosis model. LBD4WS localizes the service faults and explains the faulty reason by comparing the actual execution trace with the correct execution trace. We provide a case study to illustrate the diagnosis process of our method and use three BPEL (Business Process Execution Language) processes to conduct the experiments for evaluating our method. The experimental results show that our method is more effective than two existing MBD methods in the diagnostic accuracy for web services.