Debug Like a Scientist!
summary
Shows a repeatable workflow: observe surprising failures, form falsifiable hypotheses, run tiny experiments, and iterate. Examples: an intermittent integration test traced to a race where a non-thread‑safe before/after hook was toggling config; a CI flake fixed by removing noisy logging that triggered a Docker Compose bug; production 502s every 30 minutes traced to a 16 GB bootsnap cache slowing Passenger startup and resolved by invalidating the cache. Emphasizes simple, cheap experiments and shared notes to scale team debugging.
AI-written summary · about
chapters (7)
- 0:00The daily struggle of bugs
- 5:15The scientific approach
- 10:01Case study: Race condition
- 17:14Case study: Infrastructure issues
- 23:40Case study: Production incidents
- 32:24Scaling and coaching
- 38:40Retaining knowledge
related talks
25mDebug like a scientist
33mGetting Unstuck: Using the Scientific Method for Debugging
36mFixing Flaky Tests Like a Detective
46mBlink and You’ll Miss It: Stories From Discovering Most Annoying Bugs and Test Flakes
29mFilling the Knowledge Gap: Debugging Edition
37mDebugging Without Borders
24mHow to solve intermittent issues, aka Random Failures
01h 15mTaking Control Of Your Test Environment