The Software Test Automation System Based On Cloud Computing And CDN

As an important role and the future orientation of software testing, test automation attracts more and more attention from the IT companies and test individuals. The paper analyzed the current automation system status and pointed out what they need to improve first, then proposed the new test automation system based on cloud computing and CDN. The system requirement was analyzed and finally several key modules were designed and implemented.The testing system has several highlights when compared to current automation testing management systems:1. The system was created on could platform. It integrated the IAAS, PAAS and SAAS service model of cloud computing, the pros of different service models were fully utilized and their cons are avoided.2. The system integrated CDN with cloud to cache the static content. When CDN co-working with Windows Azure world wide data centers, the request from customers in different regions are quickly responded, thus the overall system performance are greatly improved.3. The system created rich logging system and knowledge base system as the supplement to test case execution. Rich logging system provides enough context when test case failed. Knowledge base system recorded the experience of senior testers, by reading that junior testers improved quickly.4. The system created big data based data mining system to train model and then perform prediction based on customer logging data. The service efficiency is improved and customer are more satisfied.5. The system created smart virtual machine allocation system. The test execution plan was generated by matching the test case requirement with machine condition and test cases are able to be executed in parallel. The test case execution speed was improved and the budget on virtual machines is controlled flexibly.The cloud and CDN based test automation provides testers a test automation platform with rich function and excellent performance. The test efficiency and software quality was greatly improved and testing budget was saved.