RabbitMQ — Junior¶

At junior level, focus on this question:

How do you actually declare an exchange, bind a queue, and publish/ consume a message in RabbitMQ?

The three-step setup¶

import pika

connection = pika.BlockingConnection(pika.ConnectionParameters("localhost"))
channel = connection.channel()

# 1. Declare an exchange
channel.exchange_declare(exchange="orders", exchange_type="direct")

# 2. Declare a queue and BIND it to the exchange
channel.queue_declare(queue="order_processing")
channel.queue_bind(exchange="orders", queue="order_processing", routing_key="new_order")

# 3. Publish
channel.basic_publish(exchange="orders", routing_key="new_order", body="order_data")

# Consume
def callback(ch, method, properties, body):
    print(f"Received: {body}")
    ch.basic_ack(delivery_tag=method.delivery_tag)

channel.basic_consume(queue="order_processing", on_message_callback=callback)
channel.start_consuming()

flowchart LR Declare["1. exchange_declare"] --> Bind["2. queue_declare +\nqueue_bind (routing_key)"] --> Publish["3. basic_publish"] --> Consume["4. basic_consume +\nbasic_ack"]

This is the concrete implementation of the AMQP model from Message Queues — professional: the exchange is where you publish, the binding is the routing rule, the queue is where the consumer actually reads from, and basic_ack is the explicit acknowledgment (recall the at-least-once discipline from Delivery Guarantees — ack after processing, as shown here, not before).

🎓 Takeaway: RabbitMQ's API directly mirrors AMQP's exchange- binding-queue model — nothing here is RabbitMQ-specific magic, it's a concrete implementation of the protocol concepts covered in the Message Queues topic.

Test yourself¶

What would happen if you called basic_publish to an exchange with a routing key that no queue is bound to match?
Why does the callback call basic_ack at the end, rather than at the start of processing?
What's the difference between queue_declare and queue_bind — why are these two separate steps?

Continue to middle.md.