From Hours to Minutes: Run Thousands of CI Tests in Just Minute
summary
Claims a working tool called RunRight that uses Playwright, GitHub Actions matrices, a custom state.json reporter and pre/post-commit hooks to measure per-test runtime and compute capacity. Autoscaling math uses total test-time ÷ (cores/2) ÷ target runtime to size runners and assign exact tests. Demo numbers: 35 tests in ~2 min, 105 tests in ~2.1 min, and a 3,000-test run reported at ~90s. End result: consolidated blobs produce a single HTML report; solution runs and scales as shown.
AI-written summary · about
chapters (9)
- 0:00Welcome
- 1:51Introduction
- 3:37RunRight solution features
- 8:52Slow CI system tests
- 13:25Fixing CI test times
- 16:34RunRight math and architecture
- 22:16Live demo in GitHub Actions
- 25:29End-to-end setup
- 31:26Q&A
related talks
58mSpeed up your test suite by throwing computers at it
24mSupercharging CI: Test Suite Optimization for Speed, Stability, and Savings
43mInfinite Improbability Testing- Execute All Tests in Parallel
35mMassively Continuous Integration: From 3 days to 30 minutes
20m20 Nuggets in 20 Minutes - Get more from your CI
40mInfinite improbability testing – execute all tests in parallel
43mLeveraging GitHub Copilot to 3X Your Playwright Test Efficiency
29mPlaying your tests wright