Meddelandebrokers

Meddelandebrokers are software systems that enable the exchange of messages between producers and consumers in an asynchronous, decoupled manner. They receive messages from publishers, store or route them, and deliver them to one or more subscribers or queues. By decoupling producers and consumers, meddelandebrokers allow components to operate independently, handle bursts of load, and improve reliability through queuing, retries, and failure handling.

They support common messaging models such as point-to-point (queues) and publish-subscribe (topics). In point-to-point, a message is delivered to a single consumer; in publish-subscribe, messages are delivered to multiple subscribers. Key features typically include message persistence, delivery guarantees (for example at-least-once, at-most-once, or exactly-once), routing and filtering, message transformation, dead-letter queues, retry with backoff, security controls, and monitoring and management interfaces.

Architectures vary from centralized brokers to distributed clusters that provide fault tolerance and horizontal scalability. A broker may expose protocol adapters and APIs for standard interfaces (for example AMQP, MQTT, STOMP, JMS) and define message formats, headers, and routing rules. Some systems also offer stream-oriented capabilities, ordering guarantees, and time-triggered processing.

Common implementations and ecosystems include RabbitMQ, Apache ActiveMQ, IBM MQ, Microsoft Azure Service Bus, and Google Cloud Pub/Sub. Apache Kafka is often described as an event streaming platform rather than a traditional broker, emphasizing high-throughput log-based processing, though it also functions as a broker in many architectures.

Typical use cases involve microservices communication, integration layers, asynchronous task queues, IoT data ingestion, and enterprise application integration, where decoupling and reliable delivery are desirable.