Creating event-driven microservices: the why, how and what
This talk discusses the benefits of using an event-driven architecture for microservices and how Apache Kafka can facilitate this approach. An event-driven architecture involves using events as the system’s backbone, with microservices communicating with each other by producing and consuming events. This can be more efficient than REST APIs, as it allows for asynchronous communication and reduces latency.
Several patterns can be used in event-driven architectures, such as event sourcing, CQRS (Command Query Responsibility Segregation), and event-driven APIs. Each of these patterns has its advantages and disadvantages, and the best choice will depend on the system’s specific needs.
Apache Kafka is a popular choice for event-driven microservices because it is a scalable, durable, and fault-tolerant platform for processing data streams in real-time. It can handle high throughput with low latency, making it well-suited for microservice architectures.
Teams transitioning from monoliths to microservices face complexity. Building low-latency microservices that only use REST APIs is tricky, as high latency is common. This talk covers using events as a foundation for microservice architecture, designing and planning, patterns in event-driven architectures, and why Apache Kafka is a good choice.