Regression testing is a suite of test cases that were run in the past and were successful in testing a total system or set of components. Many times, when a bug is fixed, a new test case is created to reveal the bug and to test the changes that were supposed to correct the issue.
Usually, this new test case is added to the regression testing suite to ensure that a new bug is not introduced back in software.
Regression testing can be used not only to discover the “correctness” of the software, but also to track if quality is improving.
Integration of Twister with a continuous integration system provides a powerful combination to carry out regression testing. In this configuration, the suite of test cases can be executed every time a new code change is introduced to the software and any adverse issue introduced is discovered in very short time leading to lower costs to maintain the software and improved quality.
Over the time, the number of test cases included in a regression suite can become very large and the time needed to complete the execution becomes quite lengthy. Twister can manage this scenario by execution test cases, in parallel, on distinct test beds that are configured in the same way, thus shortening time to completion and allowing large amount of test cases to be executed in a reasonable amount of time.