Research And Design Of Distributed Automation Testing Platform

Either tradition enterprise software industries or Internet industries pays more attentions on the reducing the software delivery lifecycle and improving the software quality under the fierce market completions. And with the development of software testing and test automation technologies, lots of outstanding tools or frameworks for test have been applied onto all the levels and types of testing including unit level or performance testing. At the same time, the better efficiency and stability of automation test cases execution comes to be one of the major problems limiting the continuous application of automation test technologies. The failing of costs on the computer hardware equipment entitles the distributed computing to be developed with maturity and widely adopted. The author binding the multiple disturbed computing technologies including disturbed job dispatching, disturbed computing cluster, disturbed computing cluster resource management, disturbed computing resource monitoring, disturbed storage and MapReduce programming models into the test automation to not only resolved the problems which have been described above, but also provide a generic platform to test upon Windows, Linux, Mobile operation systems supported functional and non-functional test.In order to show a clear picture of the disturbed test frameworks, this article start with the current status and problems which limited the test automation both on execution speed and execution stability and discuss the necessarily for develop a new framework to support disturbed test automation. Base on the analysis, the key components of disturbed technologies will be separately described and introduced including the disturbed job dispatching, computing cluster resource management, computing cluster resource monitor, distributed storage and MapReduce programming model. Base on the distributed computing technologies, an innovated disturbed test frameworks whose architecture implementations will be described in a detailed manner. And after all, a data analysis for the platform will be presented to prove feasibility of this disturbed test framework.