Growing Pains - What happens when your product gets BIG?
summary
Product starts in 2012 with a handful of devs and ~10k tests per day and grows to over ten times the developers with 250 builds and 20 million tests per day, shifts releases from yearly to monthly, and cuts developer friction by moving source and docs to GitHub. Each CI run compiles and runs ~70,000 tests (sequentially would be ~60 hours; average wall time ~8 hours), introduces personal builds, a build monitoring system (big‑data/AI + rotating human monitors), selective retries and comment‑triggered builds, but still battles flaky infra, long runtimes and functional silos.
AI-written summary · about
chapters (13)
- 0:00Introduction to WebSphere
- 1:23Early growth challenges
- 3:49Open Liberty open source shift
- 5:37Developer experience struggles
- 7:45Documentation and tooling improvements
- 9:12Code delivery pipeline explained
- 10:38Personal builds and build breaks
- 13:27Build monitoring and automation
- 17:45Personal build pain points
- 21:37Myths and failed practices
- 23:46DevOps team role and tools
- 25:29Current challenges and future work
- 26:53Closing remarks and Q&A
related talks
47mHow to win in software development with tighter feedback loops
46mThe Software Factory
30mThe Bitter and the Sweet of Running a Planet-Scale Build &...
20mScaling up a Super-App
01h 01mA team's transition to Continuous Delivery
28mKeeping Developers Happy with a Fast CI
46mDeveloper Productivity at Scale
38mScaling Automated Testing & Architecture from Startup to Enterprise