Six small steps towards continuous delivery
summary
Six concrete practices: start with portable checks so developers run fast local tests and simulators pre-integration; produce a golden artifact from a repeatable build with pinned dependencies; control environments with infrastructure as code and create specific disposable test environments instead of endless UATs; practice rollback and measure the four DORA metrics; split deploy from release with feature flags and incremental infra rollouts (Cloudflare 2019 regexp rollout consumed CPU and took ~30 minutes, losing ~80% capacity); extend teams to build it run it (on-call for production).
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:02Why continuous delivery matters
- 3:36Continuous delivery terms
- 6:06Portable checks and flow state
- 10:57Baked brick repeatable builds
- 13:25Environment control and testing
- 19:44Rollback and reliability
- 24:40Separating deploy and release
- 29:11Production intimacy and observability
- 33:25Summary
- 34:48Q&A
related talks
41mContinuous delivery in 10 easy mistakes
30mOur Continuous Delivery Journey
40mContinuous Delivery Anti-Patterns
23mHow to Build and Deploy Services in the Cloud fireside chat
51mBetter software, faster: principles of Continuous Delivery and DevOps
55mContinuously Delivering Continuous Delivery
41mAll Things Continuous
40mSustainable Continuous Delivery: Patterns for rapid and reliable Releases