To obtain a healthy integrated production system that achieves defined quality goals in service oriented architecture (SOA), such as availability and performance, the timely detection and resolution of failures is needed. The goal of this thesis identify the primary or root causes (faults) of a set of observed symptoms in an integrated production system that indicate degradation and failure in system components leading to abnormal system performance. Our hypothesis is, if we combine more diagnosis methods (such as more symptom repository, dynamic analysis capabilities, different artificial intelligent (AI) reasoning capabilities or other available technologies and tools) into an existing tool such as open source AspectJ based diagnostic tool Glassbox, we can find more amount of root causes and more precise --"actual" root causes. We are building an experiment platform to validate this hypothesis.