Automated Performance Testing With WebDriver
summary
Demo uses WebDriver plus Chrome DevTools to capture Lighthouse and trace metrics, and runs on Sauce Labs via an open tool Speedo. A contrived 3.5s JavaScript loop raises first paint to about 6.6s and drops the Lighthouse score from ~0.92 to ~0.53 while sent script bytes increase, proving assertions, trace export and file-level blame work. Adds jank detection (FPS, long tasks), shows how to baseline results for CI/CD, and warns that stable identical VMs are required for reliable performance tests.
AI-written summary · about
chapters (13)
- 0:00Start
- 1:42Next Billion Users
- 7:42Speed Index
- 12:15Trace Events
- 19:00Mining Bitcoin
- 20:39Caveats
- 22:57Speedo
- 24:47Demo
- 27:28Lighthouse Performance Score
- 27:54Performance Dashboard
- 35:59Recap
- 36:30Best Practices
- 37:02Separate Your Functional from Your Performance Test
related talks
31mAutomated Performance Testing With WebDriver
37mAutomated Performance Testing With WebDriver
25mAutomated Performance Testing With WebDriver
21mThe Future of Performance Tooling
14mBasics of Performance, Security, and Accessibility Testing
27mWays of Frontend performance auditing with Lighthouse
29mPerfecting Perf: Rapid, Rigorous and Reproducible Performance Experiments
40mAutomated Performance Testing In Preproduction