Research And Implementation Of Test Case Prioritization And Automated Testing Framework In Cloud Network Environment

In recent years,with the development of cloud computing and software-defined networks,more and more scholars are studying cloud-network integration and related technologies.At the same time,the scale of software testing in the cloud-network environment has become larger and the testing content has become more complex.First of all,to locate the software system’s bug faster in a short time,it is necessary to select the test case to be executed preferentially from a large number of test case sets.As a classic optimization algorithm,genetic algorithm can play a good role in the problem of test case preference,but the optimization ability of genetic algorithm is greatly affected by its parameters.Therefore,reinforcement learning method is used to optimize its crossover rate parameters,so as to improve the performance of the algorithm.Secondly,Open Stack and Open Daylight cloud-network integration platform is taken as the test object,and Selenium and Python unit test framework Unittest is combined to design and implement a corresponding automated test framework to improve the efficiency of regression testing.The main contents of this thesis are as follows:(1)In order to improve the regression testing efficiency of the cloud-network integration platform,a test case prioritization method is proposed based on reinforcement learning and genetic algorithm.Firstly,the population initial value and the selection operation of the classical genetic algorithm are improved.In this thesis,the solution of the ant colony algorithm is used as part of the initial population of the genetic algorithm,and the "elite retention strategy" is used in the selection process,which solves the problem that the classical genetic algorithm is easy to fall into the local optimal solution.Secondly,in order to obtain the crossover probability parameter that is more suitable for the population evolution in the algorithm,the reinforcement learning method is used to improve the crossover probability parameter in the classical genetic algorithm.Finally,the improved algorithm is applied to the test of the cloud-network integration platform,and the optimization capabilities of the classical genetic algorithm,the ant colony genetic algorithm and the ant colony genetic algorithm based on reinforcement learning are compared and analyzed.(2)In automated testing,there are some problems such as high coupling and low cohesion in the test case scripts written by testers,which lead to low reuse rate of test case scripts and increase the difficulty of maintenance.Therefore,after analyzing and summarizing the requirements of the cloud-network integration platform,an automated testing framework based on the Selenium test tool that conforms to the cloud-network integration environment is designed and implemented.The research results show that the ant colony genetic algorithm based on reinforcement learning proposed in this thesis has better optimization performance on the problem of prioritizing test cases.The automated testing framework is designed based on Page Object and data-driven design patterns,and has the advantages of low coupling,high reusability,and the ability to generate HTML test reports,which can effectively reduce testing costs.