--- title: "Getting started with event-driven systems" url: "https://www.bekk.christmas/post/2022/11/getting-started-with-event-driven-systems" authors: ["Ole Kristian Pedersen"] published: "2022-12-11" updated: "2024-11-07T13:05:28Z" language: "en-US" type: "article" tags: ["Sky"] keywords: ["cloud", "event-driven", "architecture", "event sourcing", "event streaming"] --- # Getting started with event-driven systems During the last year or so, I've worked a more with event-driven systems. I had to learn a lot of new terms and architecture patterns, so I turned to the internet to help me out. Along the way I've found many articles that I found useful. In this blogpost I will share some of my favorites. ## In short This article collects beginner-friendly resources for understanding event-driven systems, from core terms like event notification, event sourcing and CQRS to related architecture patterns. It also covers how event streaming differs from event sourcing and how to handle eventual consistency, duplicate events, idempotency, ordering and delivery guarantees. - **For:** Developers getting started with event-driven systems - **Good to know beforehand:** Basic knowledge of software architecture _AI-generated_ ## Intro to terms in the event-driven domain [Martin Fowler's article _What do you mean by "Event-Driven"?_](https://martinfowler.com/articles/201701-event-driven.htmll) is a nice intro to terms used when working with event-driven system. It defines "event notification", "event-carried state transfer", "event-sourcing" and "CQRS". This article also contains links to more in depth articles, and also a further reading section, so make sure to check those out! ## Starting out with event sourcing Event sourcing is not something you have to do when creating event-driven systems, but it is something that you should know about. It is an interesting architecture pattern that can have great benefits, but also be quite complex. [EventStoreDB's intro to Event Sourcing](https://www.eventstore.com/event-sourcing) is an easy-to-read, high-level introduction to event sourcing. It also introduces other related concepts (such as domain-driven design) that are nice to know. Be aware that the article use some terminology specific to EventStoreDB and introduce the "EventStoreDB way" of doing event sourcing. It's still a very nice article if event sourcing is new to you, though. [_Martin Fowler's Event Sourcing_](https://martinfowler.com/eaaDev/EventSourcing.html) is a longer, very interesting read that goes in depth on event sourcing that I recommend reading. I've read this one many times over the last year, and I learn something new every time. Like the previous article from Martin Fowler, make sure to check out the links to related articles. ## Event streaming is different from event sourcing [This article](https://event-driven.io/en/event_streaming_is_not_event_sourcing/) explains the differences between event sourcing and event streaming. It is a reply to the article [_Event sourcing is hard_](https://chriskiehl.com/article/event-sourcing-is-hard), which presents a mostly critical view of architectures using event sourcing. The latter article has some valid points, and even though it is conflating event sourcing and event streaming, it is worth a read. Event streaming usually includes some form of message broker that systems can write to, and read from. Many message brokers exist and they do have different properties, different names for the same concepts. [_Gently down the stream_](http://www.gentlydownthe.stream/) is an introduction to Apache Kafka, styled as a children's book. It is general enough to be useful even if you're using another message broker. ## Dealing with eventual consistency From the CAP theorem we know that we can't have consistency, availability and partition-tolerance at the same time, but have to choose two. In many event-driven systems, that means accepting that systems are eventually consistent (because availability and partition-tolerance are usually more important). (Micro-)services benefits from being decoupled, but have to deal with these consistency issues when state changes are transmitted as events. Cockroach Labs' blog explains how to work with [duplicate events, idempotency and out-of-order events](https://www.cockroachlabs.com/blog/idempotency-and-ordering-in-event-driven-systems/). [event-driven.io](https://event-driven.io/en/) has many good articles related to event-driven systems, but [this article about outboxes, inboxes and delivery guarantees](https://event-driven.io/en/outbox_inbox_patterns_and_delivery_guarantees_explained/) is especially interesting.