Scaling Test Automation Using A Metadata-Centric Architecture
summary
Claims a metadata hub for ETL test orchestration that replaces many copied tests with one generic routine, using SSIS + SQL Server + MSTest and object mapping of relational metadata. A single test iterates ordered inputs and validations across 100 packages, enabling parallel groups, tied test results to functionality, and cutting test code by about 10x reduction while drastically lowering false fails and execution time. The metadata model also improves BA/dev/QA alignment and lets tests be computed from changed artifacts.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 3:12Test automation in agile
- 10:19Scale and scalability
- 17:30Project size challenges
- 21:47Combinatorial test explosion
- 26:02What is metadata and why it matters
- 29:56Metadata collaboration hub
- 33:31Metadata-driven design
- 35:38Generic test case replacement
- 38:33Observations and benefits
- 40:40Future possibilities
related talks
31mHow to fix Test Automation to make DevOps faster
21mScaling Test Automation: Building Reusability for 10x Speed and Efficiency
15mTransforming Testing: The Power of TestOps
23mMetamorphic Testing
40mScaling testing & test automation without losing control
39mWhat Can Go Wrong With Testing an ETL Pipeline?
39mModern Data Pipelines Testing Techniques: A Visual Guide
24mOrchestrating automation tests for Continuous testing