Event-driven architectures are on the rise. They promise both better decoupling of components by using an event bus and improved scalability in terms of throughput. Decoupled modules help to scale your software development efforts itself. Event streaming promises to handle ever-growing amounts of „data in motion“ in real-time, event sourcing allows us to time travel, and domain events have turned out to be powerful building blocks that lead to a better understanding of underlying business requirements.
But there are also pitfalls that you’d better be aware of. For example event-notifications used inappropriately can lead to tighter coupling or cyclic dependencies between components. It is also easy to lose sight of flows across service boundaries, making it hard to understand how core business logic is actually implemented. This can get even worse if you lack tooling to get insights into your event flows. Last but not least, the event-driven approach is not well-understood by most developers or business analysts, making it hard for companies to adopt. In this talk, I will quickly go over the concepts, the advantages, and the pitfalls of event-driven utopia. Whenever possible, I will share real-life stories or point to source code examples.