Evolutionary Architectures
summary
Argues that an architecture’s target becomes a set of fitness functions that are monitored and enforced, not a fixed diagram, and shows practical techniques: continuous delivery as a foundational requirement, contract testing to let teams evolve independently, database refactoring patterns (atomic migrations) to shrink large migrations into manageable steps, and GitHub’s Scientist plus Netflix’s Simian Army for safe experiments and runtime checks. Gives a concrete surprise—87 reports blocking an upgrade—and shows these approaches actually prevent stealth breakage and enable guided incremental change.
AI-written summary · about
chapters (21)
- 0:00Why care about evolution
- 3:31Defining evolutionary architecture
- 8:05Fitness functions explained
- 13:27Types of fitness functions
- 15:10Incremental change strategies
- 16:57Continuous delivery foundation
- 17:40Optimizing multiple dimensions
- 21:55Principles of evolvable systems
- 23:44Architecting for evolvability
- 27:01Pel's law and system design
- 29:12Testing for evolvability
- 31:00Conway's law in practice
- 33:32Key techniques overview
- 34:43Database refactoring
- 37:40Orchestration vs choreography
- 41:16Contract testing benefits
- 44:14Continuous delivery importance
- 47:09Experimentation culture
- 50:19Governance with fitness functions
- 53:51Brownfield system evolution
- 58:28Revisiting fitness functions
related talks
42mBuilding Evolutionary Architectures: Principles and Practices
53mEvolutionary Architecture: Principles and Patterns
33mBuilding Evolutionary Architectures
01h 03mEvolutionary Software Architectures
39mHow will Evolutionary Architecture evolve?
35mPrinciples of Evolutionary Architecture
49mThe Principles of Evolutionary Architecture
39mPrinciples of Evolutionary Architecture