Continuous Delivery Sounds Great But It Won’t Work Here
summary
Argues Continuous Delivery works even in regulated and legacy environments: cites deployment every 11.6 seconds at scale, AWS/govcloud compliance and cloud.gov cutting 325 security controls down so only ~15–41 remain per app, and a PaaS strategy to shift controls to infrastructure. Shows engineering results: hundreds of thousands LOC churn, ~100 trunk check‑ins/day, 10–15 good builds, massive test suites (30,000+ hours) and an 8–10x productivity lift. Demonstrates test tooling like concordion driving VT on mainframes and the andon pattern for stopping the line.
AI-written summary · about
chapters (11)
- 0:00Intro
- 0:41what is continuous delivery?
- 2:26Amazon May Deployment Stats
- 4:21deployment pipeline
- 7:00compliance
- 7:55hp laserjet firmware
- 14:51part the third
- 19:46strangler application
- 21:35part the fourth
- 28:05changing culture
- 28:47TOYODA AUTOMATIC
related talks
33mContinuous Delivery - What does good really look like?
48mContinuous Delivery: Tales from Windowsland
30mOur Continuous Delivery Journey
24mWhy, What, and How of Continuous Delivery
55mThe Rationale for Continuous Delivery
59mThe continuous culture
30mModern Continuous Delivery
55mContinuously Delivering Continuous Delivery