CTL model checking in the cloud using MapReduce

The recent extensive availability of 'big data' platforms calls for a widespread adoption by the formal verification community. Cloud computing platforms represent a great opportunity to run massively parallel jobs, yet classical formal verification tools/techniques must undergo a deep technological transformation in order to exploit the new available architectures. This has raised an increasing interest in deploying verification techniques on parallel/distributed frameworks. In this paper we introduce a framework to ease the adoption of a distributed approach to verification of Computation Tree Logic (CTL) formulas on very large state spaces. The approach exploits/integrates a recently developed, parametric state-space builder. The whole framework adopts M AP R EDUCE as core computational model, and can be tailored to different modelling formalisms. The outcomes of several tests performed on (Petri-nets based) benchmark specifications are presented, thus showing the convenience of the proposed approach.