Applying CloudEvents to Level up Application Communications
summary
Minder evolves from synchronous webhooks to an async engine: started with Go channels, spent ~a year on Watermill with Postgres (durable but middling performance, Go-only, schema contention when two apps share the DB, missing CloudEvents-style metadata, dead‑letter handled in middleware causing duplicate retries), and is migrating to CloudEvents + NATS/JetStream to enable multi-language consumers, tracing and standard headers. A dual-publisher flag bridges old and new so migration proceeds safely without losing messages.
AI-written summary · about
chapters (12)
- 0:00Minder project intro
- 1:27Supply chain security
- 3:56Sync vs async messaging
- 7:47Watermill limitations
- 9:10Async workflow overview
- 13:45GitHub webhook handling
- 15:52Reminder component
- 16:56Watermill problems
- 18:41CloudEvents intro
- 21:30Transition to CloudEvents
- 23:59Why CloudEvents recommended
- 25:03Closing remarks
related talks
33mCloudEvents And Beyond!
01h 19mCloudEvents - Intro, Deep-Dive and More!
40mThinking Cloud Native, CloudEvents Future
22marmonizing Asynchronous Systems: NATS, CloudEvents and AsyncAPI.
41mOpen Standards for Building Event-Driven Applications in the Cloud
29mAPIs and Event Streaming Dualty
21mCloudEvents - v1.0 and Beyond - Discovery/Subscriptions
57mManaging Asynchronous APIs