Acceptance Testing Optimization Method for Continuous Delivery

From agile to DevOps, development methods have been extended from continuous integration to continuous delivery. In the acceptance testing before software delivery, the constantly changing software configuration, real user environment, and user-led testing, all of these extend the delivery time and increase the risk of testing. How to further optimize the test and shorten the delivery cycle are problem that must be considered to realize the end-to-end value flow in DevOps. In the testing optimization research, the methods are mainly for regression testing. There are few researches on acceptance testing, and the methods for the acceptance testing mainly focus on specific scenario, that almost none research consider how to shorten the continuous delivery cycle. Aiming at the characteristics of acceptance testing, this paper proposes an acceptance testing optimization method for continuous delivery. For different levels of test cases, test case selection and prioritization are used to optimize test cases. Firstly, the test suite related to requirements is constructed, and the test cases are selected according to the requirements. Then, the test suites are divided into two levels to prioritize. During the test execution, the use case execution actions are streamlined to shorten execution time of the acceptance testing, meet user needs as soon as possible and achieve rapid value delivery. Finally, the method is applied to actual industrial projects for experiments. The results show that the method can reduce the scale of test cases, shorten the test execution times and improve the efficiency of demand satisfaction.