Subu Iyer, Cambia Health Solutions, Inc.
Most companies today deal with lots of data. Typically, this data is in several databases and across multiple applications. Companies also have to work with data from vendors and clients. To integrate data from multiple sources, Extract Transform and Load (ETL) systems are used. These ETL projects are handled by Solution Delivery teams where each project has its own specific requirements and constraints. This makes it very hard for teams to build automated tests and processes around the ETL solutions that would help them consistently deliver high quality solutions on time and on budget. These and a set of outdated tools and practices make testing ETL jobs a time consuming and labor intensive manual process. The lack of automated testing also makes it very difficult to build and deliver incremental changes and this hurts productivity.
This paper describes how our team, the Quality Engineering and Specialized Testing (QuEST) team, is driving a change in the way teams build and test ETL jobs at Cambia Health Solutions. We are doing so by engaging with the Solution Delivery teams, setting up automated tests and providing teams hands-on training so they can continue writing automated tests. This paper will also talk about the benefits of building and growing the test framework organically in an incremental fashion. We’ll see how our team was able to overcome initial challenges and improve our ETL testing.
Target audience: Intermediate
Subu Iyer, 2014 Technical Paper, Paper, Slides
