Continuous Delivery Pipelines: How to Build Better Software Faster
summary
Deployment pipelines are the definitive statement of releasability and must give feedback under five minutes for developer checks and aim for commit-to-releasable in under an hour for the full pipeline. Evidence: State of DevOps surveys (~45k responses) show top performers are hundreds to thousands of times better on lead time, frequency and failure rate, and an industrial case cut a 9.5‑hour build to ~12 minutes commit and ~40 minutes full pipeline. Recommended practices: trunk-based development, infrastructure‑as‑code, executable specifications/BDD, TDD and automated acceptance runs.
AI-written summary · about
chapters (14)
- 0:00Intro
- 0:33Continuous Delivery
- 3:42The nature of the problem
- 7:45Best way to reduce gaps
- 9:14What determines releasability?
- 12:32Confident to make progress
- 14:13Optimize for learning
- 15:52Managing complexity
- 19:02Deployment pipeline
- 21:05Commit stage
- 24:38Acceptance stage
- 26:25Optimize for speed
- 31:38Better software faster
- 32:53Outro
related talks
01h 26mOptimising Continuous Delivery
01h 01mA team's transition to Continuous Delivery
33mContinuous Delivery - What does good really look like?
40mSustainable Continuous Delivery: Patterns for rapid and reliable Releases
51mBetter software, faster: principles of Continuous Delivery and DevOps
55mThe Rationale for Continuous Delivery
17mCI/CD - Build/Deploy Automation
24mCI/CD Best Practices to Deliver Software Faster