The increasing number of internetworking communication networks and technologies require more and more flexible computing solutions, connecting resources in a transparent, open and reliable way.

Parallelism and redundancy enables distributed service centers to provide sustained high Quality of Service and performance. Similar approaches are also used in other distributed systems. Unfortunately, the resulting concurrencies of such systems complicate control and maintenance.

In this paper, the authors present the system framework of a distributed service center and a rule-based software for the evaluation of status reports coming from the different components of the distributed service center. This software has been developed for a commercial SMS Center. Trouble shooting in the resulting distributed hardware and software system is achieved by using data mining techniques.