Complex Event Processing for Monitoring Business Processes

With Business Processes becoming more and more popular and the applications growing more complex, the need of monitoring and controlling enterprise business process systems grows. The existing solutions do not support modelbased compliance control [16] and often have limited event-abstraction capabilities. In this thesis we develop and implement a monitoring system for BPEL processes based on the ideas from Holmes et al. [15, 14], that can be used for monitoring business processes, compliance-constraint monitoring, as well as for debugging. It provides different user-interfaces for different users like high level views for non-technical personnel or debug-views for software engineers. By using model-driven engineering [5] and complex event processing we are able to build a monitoring system that can easily adapted for a variety of business process engines. From our event meta-model we generate a system, that is split into three independent components which can be distributed a long different servers and are connected via message queues. While the first component (event-listener) collects the events from the event-sources, the second component (event-processor) extract business relevant information like e.g. constraint violations from the event-streams generated by the event-listener. The third component (monitoring application) provides a user-interface for reviewing the data extracted by the event-processor.