The Key Technologies Of Software Parallel Test Based On Cloud Computing

Software testing is a labor-intensive and time-consume process. The automated test is an effective means to improve the efficiency of software testing and Cloud computing with ultra-large-scale virtualization, high availability, versatility, scalability and on-demand service characteristics, offers the possibility to build a virtual software testing environment with convenient and low-cost software testing services.This topic is about the key technologies of software parallel test based on cloud computing. Its focus on the functional automated testing research. To describe the implementation of the relationship between the software testing function points, test scene flow graph and parallel test path is defined; Through test semantic description of test cases and use it to generate the test scene flow graph and parallel test path; The parallel path will be tested using Map/Reduce mechanism on cloud testing platform; The test path will be mapped on cloud virtual machines to parallel test and the test results will be reduced. In this process, it mainly related to the operation system image matching、resource monitoring and resource selecting. After completion of the test, the test results for each test path will be integrated and then the final test report will be generated.Secondly, the AT-Cloud is set up based on the open source cloud computing framework-Eucalyptus, the automated testing framework-Autotest and the Test_Manager which is developed using Django framework. The Eucalyptus used to manage virtual machines, such as dynamic allocation, delete virtual machine and archive the flexible management of resources; the framework of Autotest used to archive the parallel paths test; Test_Manager mainly used to archive the user interaction and the communication between Eucalyptus and Autotest.Finally, a software instance ATM system is tested on the AT-Cloud to verify the validity of the proposed parallel test method and it will improve the efficiency of software testing and save the test resources.