Greybox Fuzzing of Distributed Systems
Ruijie Meng, George Pîrlea, Abhik Roychoudhury, Ilya Sergey
Abstract
Grey-box fuzzing is the lightweight approach of choice for finding bugs in sequential programs. It provides a balance between efficiency and effectiveness by conducting a biased random search over the domain of program inputs using a feedback function from observed test executions. For distributed system testing, however, the state-of-practice is represented today by only black-box tools that do not attempt to infer and exploit any knowledge of the system's past behaviours to guide the search for bugs.
Topics & Concepts
Fuzz testingComputer scienceExploitRandom testingBlack boxSoftware bugDomain (mathematical analysis)State (computer science)Software testingFunction (biology)Distributed computingMachine learningTest caseArtificial intelligenceProgramming languageComputer securitySoftwareMathematical analysisRegression analysisBiologyMathematicsEvolutionary biologySoftware Testing and Debugging TechniquesSoftware Reliability and Analysis ResearchSoftware System Performance and Reliability