Skip to content

Topic Exchange#

The Topic Exchange is a powerful RabbitMQ routing tool. This type of exchange sends messages to the queue in accordance with the pattern specified when they are connected to exchange and the routing_key of the message itself.

At the same time, if the queue listens to several consumers, messages will also be distributed among them (default scaling mechanism).

Example#

from faststream import FastStream, Logger
from faststream.rabbit import ExchangeType, RabbitBroker, RabbitExchange, RabbitQueue

broker = RabbitBroker()
app = FastStream(broker)

exch = RabbitExchange("exchange", auto_delete=True, type=ExchangeType.TOPIC)

queue_1 = RabbitQueue("test-queue-1", auto_delete=True, routing_key="*.info")
queue_2 = RabbitQueue("test-queue-2", auto_delete=True, routing_key="*.debug")


@broker.subscriber(queue_1, exch)
async def base_handler1(logger: Logger):
    logger.info("base_handler1")


@broker.subscriber(queue_1, exch)  # another service
async def base_handler2(logger: Logger):
    logger.info("base_handler2")


@broker.subscriber(queue_2, exch)
async def base_handler3(logger: Logger):
    logger.info("base_handler3")


@app.after_startup
async def send_messages():
    await broker.publish(routing_key="logs.info", exchange=exch)  # handlers: 1
    await broker.publish(routing_key="logs.info", exchange=exch)  # handlers: 2
    await broker.publish(routing_key="logs.info", exchange=exch)  # handlers: 1
    await broker.publish(routing_key="logs.debug", exchange=exch)  # handlers: 3

Consumer Announcement#

First, we announce our Topic exchange and several queues that will listen to it:

exch = RabbitExchange("exchange", auto_delete=True, type=ExchangeType.TOPIC)

queue_1 = RabbitQueue("test-queue-1", auto_delete=True, routing_key="*.info")
queue_2 = RabbitQueue("test-queue-2", auto_delete=True, routing_key="*.debug")

At the same time, in the routing_key of our queues, we specify the pattern of routing keys that will be processed by this queue.

Then we sign up several consumers using the advertised queues to the exchange we created:

@broker.subscriber(queue_1, exch)
async def base_handler1(logger: Logger):
    logger.info("base_handler1")


@broker.subscriber(queue_1, exch)  # another service
async def base_handler2(logger: Logger):
    logger.info("base_handler2")


@broker.subscriber(queue_2, exch)
async def base_handler3(logger: Logger):
    logger.info("base_handler3")

Note

handler1 and handler2 are subscribed to the same exchange using the same queue: within a single service, this does not make sense, since messages will come to these handlers in turn. Here we emulate the work of several consumers and load balancing between them.

Message Distribution#

Now the distribution of messages between these consumers will look like this:

await broker.publish(routing_key="logs.info", exchange=exch)  # handlers: 1

Message 1 will be sent to handler1 because it listens to "exchange" using a queue with the routing key "*.info".


await broker.publish(routing_key="logs.info", exchange=exch)  # handlers: 2

Message 2 will be sent to handler2 because it listens to "exchange" using the same queue, but handler1 is busy.


await broker.publish(routing_key="logs.info", exchange=exch)  # handlers: 1

Message 3 will be sent to handler1 again because it is currently free.


await broker.publish(routing_key="logs.debug", exchange=exch)  # handlers: 3

Message 4 will be sent to handler3 because it is the only one listening to "exchange" using a queue with the routing key "*.debug".