1. What do you think is the biggest risk in this kind of project? The complexity of the QA testing effort at HPHC, with as many as 30–40 applications or functional components going into the production environment at the same time, was significant and required a great deal of thought and planning. HPHC needed a new way of thinking about QA and the path to production (e.g., synchronizing promotions of dependent parts) of many different but interdependent applications (quite different than the software companies who develop a single product or application).

◾◾Creating and maintaining the environments for the extensive integrated testing in a SOA environment required a new approach. The costs to create a complete replica of the entire production environment for testing purposes were prohibitive. As a result, HPHC and Dell created the following:

◾◾ a team responsible for managing test environments.

◾◾ development of integrated test environments that consist of the application to be tested and its immediate neighbors—this is a continually ongoing process.

◾◾ a system-integrated test environment, which was a subset of the full production environment that would allow end-to-

end testing of a variety of scenarios. This environment included all of the applications upstream and downstream (not just the immediate neighbor applications) that were part of the end-to-

end scenarios being tested.

◾◾ a team responsible for creating test data needed in the environments. It was simply too much to expect project teams to be responsible for the collection of all the needed data, assembling it in a way consumable by the applications in a test environment, and then providing a “reset” capability for continued use. Teams provide the scenarios and descriptions of the data needed to this team, who then get it, assemble it, and proof it with the teams prior to the actual testing work.