How to control parallel or synchronous execution based on task type

[v2boardbot]

Consumer code:

async def main():
    await app.run()


def run(queue_name):
    task_queue = RabbitQueue(queue_name, auto_delete=True, routing_key='task')
    @broker.subscriber(queue=task_queue, no_ack=True)
    async def task(action, data, msg: RabbitMessage):
        print('received task', action, data, os.getpid())
        if action == 'send_message':
            sleep = random.randint(10, 15)
            await asyncio.sleep(sleep)
            print('task', action, data, 'Processing completed', datetime.now())
            await msg.ack()
        else:
            await msg.ack()
            return 'ok'

    asyncio.run(main())

If the action is send_message, I want to sleep for a while after completion and continue to execute the task of action send_message, but the tasks of other actions should be processed immediately.
broker = RabbitBroker('xxxxx', max_consumers=1)
When I set max_consumers to 1, other action tasks will also be queued for processing.

[Lancetnik]

Hi! I am not sure, what behavior you expect: max_consumers is just an option to set channel QOS (https://www.rabbitmq.com/docs/consumer-prefetch)

If you set it to 1, all of your application consumers will be working in synchronous mode. You can't mix this behavior on RMQ level.

But, you can figure out something with asyncio.Lock.