Netflix takes inspiration from Certus SFI research paper

15 November 2019

netflix-new-icon.png

Recently, Netflix created and deployed a tool that schedules the execution of test cases on a variety of devices. For that, they took inspiration from the Certus research paper “Reinforcement Learning for Automatic Test Case Prioritization and Selection in Continuous Integration”. The paper is written by Simula and Certus researchers Helge Spieker, Arnaud Gotlieb, and Dusica Marijan as well as Morten Mossige, Associate Professor at the University of Stavanger.

In this paper, the authors introduced Retecs, a new method for automatically learning test case selection and prioritization in continuous integration with the goal to minimize the round-trip time between code commits and developer feedback on failed test cases. The method exploited reinforcement learning to select and prioritize test cases according to their duration, previous last execution, and failure. By applying Retecs on data extracted from an ABB's use case, we showed that reinforcement learning was able to automatically perform adaptive test case selection and prioritization in continuous integration.

The Certus Centre for Software Verification and Validation

The Certus Centre is an 8-year research project that engages Simula Research Laboratory as the host institution with ABB Robotics, the Cancer Registry of Norway, Cisco Systems Norway, and Esito as user partners. The project is financially supported by the Research Council of Norway. The centre is led by Chief Research Scientist Arnaud Gotlieb.