Parallel Testing in Behavior Driven Development

The testing process consists of activities that demand efforts asproducing, executing, and validating test scenarios. Covering alltest scenarios manually is unfeasible since it is error-prone andlabor-expensive. Thereby, partial or complete automation reducescosts and increases tests’ effectiveness. The increasing availabilityof hardware resources provides opportunities to scale testingusing parallel execution of test cases or suites blocks. Some toolsperform parallel execution of tests, but their use requires complicatedsettings, and when combined with some methodologies asBehavior-Driven Development, it may create an overhead for users.This paper presents the Multi-Threaded Testing (MTT) tool for parallelexecution of test scenarios in the context of Behavior-DrivenDevelopment that aims to reduce the computational time requiredto test Java projects. Furthermore, the present paper reports anexperimental study to evaluate the MTT tool’s performance intwo different hardware configurations. Our results demonstrate theMTT reached a speedup of 4,59 using ten threads in CPU Intel Corei5-9300H with an efficiency of 46%, and a speedup of 3,45 with anefficiency of 43% using eight threads in CPU Intel Core i7-7700HQ.