How to Make Your Functional Tests Truly Fast
summary
The talk addresses making functional (end-to-end) tests much faster by going beyond naive parallel runs. It describes problems like tests sharing the same data, overlapping functionality, use of real databases and external services that block parallel execution, and presents practical techniques and approaches to resolve data dependencies and enable reliable parallelization of functional tests.
AI-written summary · about
chapters (18)
- 0:00Introduction
- 3:17Test independence via data
- 7:46Data isolation with accounts
- 9:49Scaling via parallelization
- 12:02Browser performance
- 16:21Direct sessions instead of browser
- 17:47Test atomicity
- 19:52Optimizing via test splitting
- 21:42Skipping non-critical steps
- 23:29Direct database control
- 25:43Integration tests and dependencies
- 27:34Isolating external dependencies
- 29:23Disabling external services
- 30:51Proxy for speed
- 33:41Proxy resource analysis
- 34:44Avoiding waits and delays
- 36:15Minimal waits and monitoring
- 37:39Performance analysis tools
related talks
41mFunctional Tests (Selenium) in Practice
40mPitfalls of Functional Testing and How to Overcome Them (with TestCafe)
44mTesting the untestable - patterns and use cases analysis
44mThe Hell of Automated Tests
43mHyper-Parallelism - The Panacea of Automated Testing?
42mMeasuring Performance with Functional Tests: A Low-Effort Approach for High Impact
43mFunctional Testing for High-Load Projects
51mAre Your Test Slowing You Down?