Test Case Prioritization for Multiple Processing Queues

Test case prioritization is an effective technique that helps to increase the rate of fault detection or code coverage in regression testing. However, all existing methods can only prioritize test cases to a single queue. Once there are two or more machines that participate in testing, all exiting techniques are not applicable any more. To extend the prioritization methods to parallel scenario, this paper defines the prioritization problem in such scenario and applies the task scheduling method to prioritization algorithms to help partitioning a test suite into multiple prioritized subsets. Besides, this paper also discusses the limitation of previous metrics and proposes a new measure of effectiveness of prioritization methods in a parallel scenario. Finally, a case study is performed to illustrate the algorithms and metrics presented in this article.