Asynchronous Microservices in Practice

You're at a restaurant. The waiter takes your order, walks to the kitchen, asks if they have burgers, waits for an answer, walks back, confirms, takes the order, walks back again, then stands next to the kitchen until your food is ready. The table in the corner is getting restless. Other orders are backing up. That waiter is blocked. We don't want that at dinner. Why build our software like it? When services are modelled around data entities instead of business capabilities, they end up doing exactly this: calling and waiting. Our code is tightly coupled and everything slows down together. Six months into a microservices project you've got something slower and harder to deploy than the monolith it replaced. The fix is a different way of thinking. If you think of your system as teams of people, they wouldn't queue up waiting for a response. They'd pass a note when they had something to say, call it out when something changed, and keep moving. In this talk we'll work through the patterns that reflect this: commands, events, and sagas, then get into running code that shows how to do it in production. If your team is thinking about building a distributed architecture, or already in one that feels wrong, this talk is for you.