Granularity on Persistent Data Flow Testing of Active Database Applications

Active databases have been traditionally used as an alternative to implement persistent data requirements of applications on several knowledge domains. Their principle is the activation of tasks with specific functionalities as a response to events. These reactive abilities are generally expressed with active rules defined within the database itself. We investigate the use of data flowbased testing to identify the presence of faults in active rules written in SQL. Our research is based on the precision of data flow analysis, also named as data flow granularity, aiming at comparing different granularities and preliminarily evaluating their fault-revealing effectiveness. This analysis has an important impact on the cost of database application testing. The results point to higher granularities do not improve the fault detecting ability.