Event-Based API Patterns and Practices
summary
Lists seven concrete design patterns for event APIs: thin event notifications with hypermedia links to fetch current state, event-carried state transfer for full snapshots (common with Apache Kafka and Pulsar), nested structured payloads to avoid name collisions, schema evolution rules to avoid breaking historical events, explicit offline/recovery support (GitHub webhook retry APIs as an example), and strict separation of internal versus external events. Recommends documenting messages with the AsyncAPI spec and choosing webhooks, SSE, WebSockets, gRPC streams or distributed logs based on delivery and consumer needs.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:13What are asynchronous APIs
- 4:59Webhooks
- 7:27Server Sent Events
- 10:00Design Patterns
- 11:30Thin Event Notification
- 12:30HyperMediaDriven Events
- 13:40EventCarryState Transfer
- 14:46Structured Event payloads
- 17:02Evolution II Event Schema
- 19:24Offline Error Recovery Support
- 21:28Separate Internal and External Events
- 23:24People are doing
- 24:22The evolving enterprise platform
related talks
03mWhy AsyncAPI Is Essential for Modern Event-Driven Architectures
23mEvent API Products: Maximizing the Value of Your Event-Driven Architecture.
38mEvent Design & Event-first Development
32mUnifying API and event driven architectures
26mAPI Management, meet Event Management
38mBest Practices to Design & Build Event-driven Applications
34mAPI Management, Meet Event Management
25mAccelerate Event Enabling the API Economy