RabbitMQ crash course (part 1)

Let’s start from starting single RabbitMQ instance im Docker container.

Steps

1. Download latest docker image

   docker pull rabbitmq
   

2. Run RabbitMQ as a container

   docker run --rm -d --hostname rabbit0.amql.local --name rabbit0 --label rabbit0 rabbitmq:latest
   

3. At this point you should have running RabbitMQ container which you should be able connect over AMQP.

   radek@ant:~$ docker ps
   CONTAINER ID        IMAGE               COMMAND                  CREATED             STATUS              PORTS                                NAMES
   c23bda22c2fe        rabbitmq:latest     "docker-entrypoint.s…"   5 minutes ago       Up 5 minutes        4369/tcp, 5671-5672/tcp, 25672/tcp   rabbit0
   

4. Check status of your RabbitMQ

   docker exec -it rabbit0 rabbitmqctl status
   docker exec -it rabbit0 rabbitmqctl -p / list_queues
   docker exec -it rabbit0 rabbitmqctl list_queues
   docker exec -it rabbit0 rabbitmqctl list_vhosts
   docker exec -it rabbit0 rabbitmqctl list_connections
   

Congrats you now have now fullty running RabbitMQ to which you can connect to.

Below Python code needs to be run on the same host like your RabbitMQ container in order to work, it implements simple producer-consumer model.

1. Install Python Pika which is AMQP client library. Running below example inside Python virtual environment is recomended.

   python3 -m venv /var/tmp/venv3-amqp
   source /var/tmp/venv3-amqp/bin/activate
   pip install -U pip
   pip install pika
   

2. Start from extracting IP address which is assigned to your RabbitMQ container and exporting it as env variable

   export RABBIT_IP=$(docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' rabbit0)
   echo $RABBIT_IP
   

3. Push some messages to the queue on your AMQP broker.

   #!/usr/bin/env python3
   """ If script executed with arguments, all of them will be send
   over AMQP to reciver:
   
    $ python producer.py arg0 arg1 arg2 ...
   
   """
   
   import pika
   import os
   import sys
   
   QNAME='queue0'
   RABBIT_PORT = 5672
   RABBIT_IP = os.environ.get("RABBIT_IP", None)
   
   def message():
       _msg = "AMQP MESSAGE"
       if len(sys.argv) > 1:
           _msg = ' '.join(sys.argv[1:])
       return _msg
   
   def main():
       if RABBIT_IP is None:
           print("Missing RABBIT_IP env variable")
   
       msg = message()
       print('AMQP producer connecting to {}:{} with message: \"{}\"'.\
             format(RABBIT_IP, RABBIT_PORT, msg))
       connection = pika.BlockingConnection(pika.ConnectionParameters(RABBIT_IP, RABBIT_PORT))
       channel = connection.channel()
       channel.queue_declare(queue=QNAME)
       channel.basic_publish(exchange='',
                             routing_key=QNAME,
                             body=msg)
       connection.close()
   
   if __name__ == '__main__':
       main()
   

   Each time you run this script it will push one message to queue0 on your RabbitMQ. You can check your queued messages on rabbit:

   radek@ant:~$ docker exec -it rabbit0  rabbitmqctl list_queues
   Timeout: 60.0 seconds ...
   Listing queues for vhost / ...
   name    messages
   queue0  5
   

   As you can see there are 5 messages in queue queue0.

4. Let’s now consume all the messages from queue.

   #!/usr/bin/env python3
   
   import pika
   import os
   
   RABBIT_IP = os.environ.get("RABBIT_IP", None)
   RABBIT_PORT = 5672
   QNAME='queue0'
   
   def callback(ch, method, properties, body):
       print(" [x] Received %r" % body)
   
   def main():
       if RABBIT_IP is None:
           print("Missing RABBIT_IP env variable")
   
       print('AMQP consumer connecting to {}:{}'.format(RABBIT_IP, RABBIT_PORT))
       connection = pika.BlockingConnection(pika.ConnectionParameters(RABBIT_IP, RABBIT_PORT))
       channel = connection.channel()
       channel.basic_consume(queue=QNAME,
           auto_ack=True,
           on_message_callback=callback)
       channel.start_consuming()
       connection.close()
   
   if __name__ == '__main__':
       main()