9. Asynchronous processing
ASGI (Asynchronous Server Gateway Interface) is a spiritual successor to WSGI, intended to provide a standard interface between async-capable Python web servers, frameworks, and applications.
Where WSGI provided a standard for synchronous Python apps, ASGI provides one for both asynchronous and synchronous apps, with a WSGI backwards-compatibility implementation and multiple servers and application frameworks.
A task queue implementation for Python web applications used to asynchronously execute work outside the HTTP request-response cycle. Celery can be used to run batch jobs in the background on a regular schedule.
You want your WSGI server to respond to incoming requests as quickly as possible.
Each request ties up a worker process until the response is finished.
Moving work off those workers by spinning up asynchronous jobs as tasks in a queue is a straightforward way to improve WSGI server response times.
9.2.2. Celery daemon
Workers that handle whatever tasks you put
Each worker will perform a task
When the task is completed will pick up the next one
The cycle will repeat continuously
Waiting idly when there are no more tasks
at time intervals (every 5 seconds or once a week),
on a specific date or time (at 5:03pm every Sunday)
pip install celery
9.2.5. Basic usage
Define task in
tasks.pyfile by decorating function
Run Celery workers with
tasksmodule (use verbose "info" logging)
celery -A tasks worker --loglevel=info
Call function asynchronously by using
.delay()special method added by Celery
from tasks import add result = add.delay(4, 4)
If you want to store results use:
app = Celery('tasks', backend='db+sqlite:///results.sqlite', broker='amqp://')
result.ready() # False result.failed() # False result.successful() # False result.state # PENDING -> STARTED -> SUCCESS # 'PENDING'
RabbitMQ is the most widely deployed open source message broker
Implementation of the Advanced Message Queuing Protocol (AQMP)
AQMP is an open standard
A message is a way of exchanging information between application, servers and processes. When two applications share data among themselves, they can decide when to react to it when they receive the data. To exchange data effectively, one application should be independent of another application. This independence part is where a message broker comes in.
- Message Broker
A message broker is an application which stores messages for an application. Whenever an application wants to send data to another application, the app publishes the message onto the message broker. The message broker then finds out which queue this message belongs to, finds out the apps which are connected to that queue and so, those apps can now consume that message.
The message broker app, like RabbitMQ, is responsible for saving that message until there is a consumer for that message. Queues are just virtually infinite buffers which store message packets.
docker run -d -p 5462:5462 rabbitmq
Ubuntu or Debian package:
echo "deb http://www.rabbitmq.com/debian/ testing main" >> /etc/apt/sources.list curl http://www.rabbitmq.com/rabbitmq-signing-key-public.asc | sudo apt-key add - sudo apt-get update sudo apt-get install -y rabbitmq-server
9.3.3. Management Console
Manage users and their permissions and roles
Create new queues
Manage queues, monitor their consumption rate etc.
Purge data which is currently on queues
Send and receive messages
Memory usage against each queue and by the overall process
sudo rabbitmq-plugins enable rabbitmq_management
Default credentials is:
sudo rabbitmqctl add_user admin password sudo rabbitmqctl set_user_tags admin administrator sudo rabbitmqctl set_permissions -p / admin ".*" ".*" ".*"
9.3.4. Manage RabbitMQ
service rabbitmq-server start
service rabbitmq-server stop
service rabbitmq-server restart
service rabbitmq-server status