Keynote: Bringing a Ferrari to a Horse Race
summary
Started with RabbitMQ as a background job runner and turns it into the event bus for an intentionally distributed system, building an internal messaging framework that enforces Google Protobufs for message contracts, uses fanout exchanges with predictable names, and creates durable queues so messages survive failures; moves from self-managed Rabbit clusters to CloudAMQP managed Rabbit cluster, runs services on AWS/Fargate with Terraform, adopts Elixir/BEAM supervision to control retries and failures, sees a few hundred to low-thousands messages/second in practice, and finds the architecture reliable and maintainable.
AI-written summary · about
chapters (6)
- 0:00Start
- 6:55Rabbit as a Background Jobs Processor
- 8:04Networks are hard
- 15:39Pitfalls of Synchronous Communication
- 18:49Pitfalls of Async Communication
- 30:19Peeling responsibilities - StudentRegistered
related talks
17mFun, interesting and common RabbitMQ use cases
36mDRb vs. RabbitMQ Showdown
44mKeynote: Erlang Powered Rabbits
57mReliability Nirvana
55mKeynote: Announcing Broadway
41mHow We Rabbit at AWeber
39mBuilding an event-driven notification system in Go and RabbitMQ
57mGame of Streams: How to Tame and Get the Most from Your Messaging Platforms