Delivering a re-usable QA framework for a complex system update.
Collaborative, automated test approach leads to successful project delivery and lays groundwork for better, more efficient QA in future projects
The client required end-to-end QA for a system update involving a third-party software service, with associated changes to in-house IT applications and databases, in systems managed by two separate IT teams. Previous revisions of the system had been tested manually, with only basic data verification. No automated testing had been implemented for the system, since no framework existed to automate tests that crossed the various server and database boundaries that the system used.
Plaster Group engaged with the client’s in-house QA as well as the SaaS provider’s QA team. A subsequent analysis of the system for testability, led to the determination that a combination of manual and automated testing would be the best approach.
Plaster Group implemented a basic test framework using .Net Microsoft technologies, which could harness the services and tasks used by the system as well as make the necessary database calls to efficiently verify the large amounts of data flowing through the system. Using carefully crafted test data, and working with the SaaS vendor as well as client QA resources to connect sandbox test systems, reliable automated tests were created which could be run repeatedly throughout the project to validate the updates and verify the system stability.
At the same time, the Plaster Group provided overall leadership and guidance to manual test planning and execution, authoring a formal test plan, coordinating the work across QA teams, and managing external status communications.
The project was completed on time at a high level of quality exceeding client expectations. An improved model for cross-team QA collaboration was established. Moreover, the test automation uncovered existing bugs not directly related to the update, which were subsequently fixed, improving the overall quality and reliability of the IT data systems involved. The test framework created for the project was used soon afterwards for a subsequent, separate project involving some of the same servers and databases, saving significant test time for that project as well – and is now in place for future projects and regression testing.
● Team Foundation Server
● Microsoft Unit Test Framework
● Microsoft Test Manager