Structural Testing of Active DataBases

Martín Balzamo
Martina Marré
Daniel Yankelevich

Departamento de Computación, FCEyN, Universidad de Buenos Aires, Argentina


Active databases (ADBs) are databases that include active components or agents that can execute actions. The rise of active databases in the picture of software development has a great impact on software systems and in the discipline of software engineering. However, we still lack the foundations that are needed to adequately support this new tool. These foundations are needed in order to properly apply known software engineering techniques to ADBs and systems that use them. Among the methods and techniques used to improve quality, we count systematic testing. In this work, we generalize structural testing techniques to ADB systems. We introduce a model of active databases, called dbgraph, suitable for testing. We show that dbgraphs can be used to generalize structural testing techniques for ADBs. Moreover, we introduce several new structural criteria aimed at find errors in a set of rules for an ADB. We also compare the strength of the coverage criteria presented in this work.

Supported in part by UBACyT grant EX186.