A Service-Oriented Alarm Correlation Method for IT Service Fault Localization
Cheng Lu, Qiu Xue-song, Nan Zhang · 2006
To improve the quality of IT services it is important to quickly and accurately detect and diagnose its faults. Because associations exist among entities, one fault in a single entity may cause a number of accompanying alarms beside the root alarm during fault propagation. Service-oriented alarm correlation analysis, that isolates the root causes or faults from numerous alarms, is an important method to ensure quality of service. This paper firstly provides an entity relationships model of IT infrastructure. Based on the fault propagation map depicted in the model, we then propose a service-oriented alarm correlation method. We have validated our efforts by developing a prototype and testing it in a real environment