Juggling with Data: On the Lack of Database Monitoring in Long-Living Software Systems

Abstract

Long-living software systems often challenge associated software engineers and operators with changing requirements or increasing workload. In addition, performance issues or customer requests may cause inevitable software updates or refactoring tasks. These legacy systems are often based on outdated technologies and are poorly documented. In combination with insufficient knowledge of the (actual) system, the necessary transformations can be difficult. While the described situation is discussed in the software engineering community, databases are rarely covered. As databases are an essential part in almost every soft- ware system, they are affected by evolutionary tasks as well and have to face the same challenges. In this paper, we (i) describe problems based on non-existing database monitoring in long-living software systems and (ii) propose an approach as a solution, which addresses the described problems, and supports developers and operators in performing evolutionary tasks alike. Finally, we conclude and delineate open questions concerning our envisioned approach.