When agile methods and automated programming accelerate a team’s delivery pace, the quality assurance function better be ready to keep up. Every data warehousing project faces similar challenges regarding quality assurance:
1) A large number of system components needing testing – typically hundreds of them
2) Teams should employ many different types of tests upon those components – over 40 at last count
3) Agile teams change their code base daily– necessitating a full regression test every night
The resulting number of test-case executions a team should perform during the length of a project can and should reach into the tens of thousands—otherwise the team will be relying upon “code and pray” for the untested portions of the application’s data-transform logic.
As hard as it is to author hundreds of test cases, when is your team going to find time to run them all every night? Agile data warehousing teams absolutely need test automation.
Ceregenics utilizes the latest test automation tools in its business analytics development projects, allowing teams to implement full, nightly regression testing in one tenth the time and a similar reduction in expense. We utilize only test engines that understand the unique aspects of DW/BI testing, including:
• subject areas, workflows, and architectural layers
• error classes such as dirty data, missing data, and incoherent sequencing
• time points such as mid-month, end-of-month, and end-of year
By combining these testing aspects, we generate thousands of test scenarios that the test engine then executes nightly against the current build of your data warehouse. Every morning, your stakeholders and project managers receive a dashboard of evidence-based status reporting for their projects, with the evidence depicted being the pass-fail results of the previous night’s tests. Moreover, your developers receive a detailed defect report to guide their next day’s worth of agile programming.