As software-driven organizations collect and analyze more customer data than ever before, ETL (Extract, Transform, Load) has become more important than ever. Meanwhile, the role of ETL has expanded with the advent of Continuous Delivery and “agile” development, as multiple teams require instant, in parallel access to the data they need, in the right state.
This paper sets out a step-by-step guide to fully automating ETL testing, making it more rigorous and more reactive to change. It considers how formal modelling can be used to eliminate defects earlier in the development lifecycle, while also automating test case and data creation, test execution, and the implementation of change. Doing so, the integrity of the ETL routines which organizations rely on is assured, allowing them to deliver quality software, on time and within budget.