Compositional Reverification of Probabilistic Safety Properties for Large-Scale Complex IT Systems

摘要

Compositional verification has long been regarded as an effective technique for extending the use of symbolic model checking to large, component-based systems. This paper explores the effectiveness of the technique for large-scale complex IT systems (LSCITS). In particular, we investigate how compositional verification can be used to reverify LSCITS safety properties efficiently after the frequent changes that characterise these systems. We identify several LSCITS change patterns-including component failure, join and choice-and propose an approach that uses assume-guarantee compositional verification to reverify probabilistic safety properties compositionally in scenarios associated with these patterns. The application of this approach is illustrated using a case study from the area of cloud computing.