RabbitMQ 架构详解

RabbitMQ is a message broker — a smart post office for your apps. Let's break down the components that make it tick.

At its core RabbitMQ is a message broker: a server that receives messages from producers, stores them reliably, and delivers them to consumers.

A producer is any app that sends messages to RabbitMQ. It doesn't know or care who will receive them, so it runs completely independently.

A consumer pulls messages from a queue and acts on them. Many consumers can share the work, processing messages in parallel.

A queue is the buffer where messages line up in order, waiting for a consumer. Make it durable and it survives broker restarts.

Producers never publish straight to a queue. They publish to an exchange, which acts like a sorting desk that routes messages onward.

Exchanges route by type: fanout broadcasts to all queues, direct matches an exact key, and topic matches wildcard patterns.

A binding is the rule linking an exchange to a queue: match this criteria, send it here. No bindings, and messages have nowhere to go.

A virtual host is a logical RabbitMQ inside the main one — its own queues, exchanges, and bindings — so teams share a server without colliding.

The full journey: a producer publishes to an exchange, bindings route it to one or more queues, and a consumer reads it from there.

Producer, exchange, binding, queue, consumer — which does which? Let's see if the components stuck.

That's the RabbitMQ map: producers send, exchanges route, bindings connect, queues store, consumers process — and vhosts keep it all isolated.