Data on the Inside, Data on the Outside
summary
Explains Pat Helland’s Data on the Inside vs Data on the Outside distinction and gives concrete implementation guidance: when to cache reference data, how to version and deduplicate messages, and how idempotence avoids duplicates. Compares transport choices with practical tradeoffs (Kafka vs RabbitMQ), shows the outbox pattern and change data capture as ways to guarantee eventual consistency, and outlines when to prefer local caches over RPC to avoid temporal and behavior coupling.
AI-written summary · about
chapters (14)
- 0:00Introduction and context
- 3:46The concept of SOA
- 7:44Principles of microservices
- 16:21Data on the inside vs outside
- 18:33Handling stale data
- 22:25Idempotence and messaging
- 26:38Request vs event architectures
- 34:51Reference data and caching
- 40:23Active services and feeds
- 43:33Event-driven state transfer
- 47:12Data sharing strategies
- 54:00Ensuring system correctness
- 58:35Service composition
- 1:01:22Sagas and error handling
related talks
49mEvent Driven Collaboration
31mData Unleashed: A Developer's Perspective on Navigating the Architecture Maze
53mData Access and Data Sharing in Microservices
01mA Change-Data-Capture use-case: designing an evergreen cache
52mData Over Code: A Short Hyperbole
47mApplication design patterns with in-memory datastores
57mEvent Driven Collaboration
55mThe data dichotomy: Rethinking data and services with streams