In order to study the distributed and automated testing of large-scale software, a distributed automated software testing platform (DASTP) was presented, which was based on the analysis of an existing software testing framework and the idea of distributed continuous software quality assurance. The prototype system was implemented. The platform integrated a series of tools required in the software testing, and could use free resources in Internet to complete the continuous integration and testing of large-scale software. A task scheduling algorithm based on part-time constraint and an improved ACO algorithm for set partitioning problem were proposed. These two algorithms can partition testing task into subtasks and then schedule these subtasks automatically. By running MySQL testing in the prototype system, the feasibility of the platform architecture and the effectiveness of the algorithms were verified.
