Progressive Delivery: Patterns & Benefits of Decoupling Deploy from Release
summary
Progressive delivery shows how to decouple deploy from release using runtime feature flags and dynamic config so changes reach production but only affect a subset of users, enabling randomized deterministic splits (hash mod 100), automated guardrails and telemetry that replace manual analysis. Examples: LinkedIn automates rollouts and fixed a bad recommendation by changing targeting in minutes, Booking.com detects and kills regressions in seconds, TVNZ cuts analysis from weeks to two weeks for reliable A/B decisions. The pattern works and reduces blast radius.
AI-written summary · about
chapters (6)
- 0:00Intro
- 1:57Progressive delivery and its benefits
- 8:08Feature flag
- 19:54Lessons learned
- 26:32Checklist to level up
- 36:03Outro
related talks
22mProgressive Delivery: The Logical Evolution CI/CD
06mDeploying Radically, Democratizing Release
40mContinuous Delivery Anti-Patterns
53mDeliver Results, Not Just Releases: Control & Observability in CD
18mKeynote: Safe at any Speed - How Progressive Delivery Eliminates Risk
01h 28mThe fantastic four coding patterns of Continuous Delivery
43mCloud-Native Progressive Delivery
16mProgressive Delivery with Observability