High-Performance Testing: Parallelizing Functional Tests for Computer Systems Using Distributed Graph Exploration

Functional testing of complex hardware and software systems has long been recognized as an immensely computerintensive task. Consisting of a huge number of interacting components, computer systems are hard to be verified due to the well-known fundamental problem – combinatorial state explosion. One of the ways to overcome the complexity is to use abstract models for generating test sequences and checking design correctness. However, models of really complex systems are complex themselves, which leads to enormously long test sequences (tests are usually targeted at covering all model states reachable from the initial one). In this paper, we suggest a method for high-performance generation and execution of modelbased tests based on the distributed exploration of a system’s graph model. The key feature of the method is that parallelization is done dynamically and fully transparently for a user. Keywords–functional testing; simulation-based verification; combinatorial state explosion; distributed computing; graph exploration; UniTESK technology