Navigate back to the homepage

Introduction to Message Queue:Build a newsletter app using Django, Celery, and RabbitMQ in 30 min

Nancy Chauhan
April 26th, 2021 · 4 min read

Messaging Queues are widely used in asynchronous systems. In a data-intensive application using queues makes sure users have a fast experience while still completing complicated tasks. For instance, you can show a progress bar in your UI while your task is being completed in the background. This allows the user to relieve themselves from waiting for a task to complete and, hence, can do other jobs during that time.

A typical request-response architecture doesn’t cut where response time is unpredictable because you have many long-running requests coming. If you are sure that your systems request will exponentially or polynomially go large, a queue could be very beneficial.

Messaging queues provide useful features such as persistence, routing, and task management. Message queues are typical ‘brokers’ that facilitate message passing by providing an interface that other services can access. This interface connects producers who create messages and the consumers who then process them.

We will build a newsletter app, where users can subscribe to various newsletters and they will receive the issues regularly on their emails. But before we proceed let’s understand the working of workers + message queues.

Message Queue

Workers & Message Queues

Workers are “background task servers”. While your web server is responding to user requests, the worker servers can process tasks in the background. These workers can be used for sending emails, making large changes in the database, processing files, etc.

Workers are assigned tasks via a message queue. For instance, consider a queue storing a lot of messages. It will be processed in a first-in, first-out (FIFO) fashion. When a worker becomes available, it takes the first task from the front of the queue and begins processing. If we have many workers, each one takes a task in order. The queue ensures that each worker only gets one task at a time and that each task is only being processed by one worker.

We will use Celery which is a task queue implementation for Python web applications used to asynchronously execute work outside the HTTP request-response cycle. We will also use RabbitMQ, which is the most widely deployed open-source message broker. It supports multiple messaging protocols.

Build Newsletter App

We will build a newsletter app where a user can subscribe to various newsletters simultaneously and will receive the issues over their emails regularly.

Product Overview

We will have our newsletter app running as a Django app with celery. Whenever authors publish a new issue the Django app will publish a message to email the issue to the subscribers using celery. Celery workers will receiver the task from the broker and start sending emails.



  • Python 3+ version
  • Pipenv

Setup Django

Create a folder newsletter locally and install Django in a virtual environment. Inside folder run:

1pipenv shell
2pipenv install django

Create an app:

django-admin startproject newsletter_site .

Setup the models:

python migrate

Make sure it works and visit :

python runserver 8000

Create the newsletter app:

python startapp newsletter


  • Install celery
  • Install dotenv for reading settings from the environment.
  • Install psycopg2-binary for connecting with Postgres.
1pipenv install celery
2pipenv install python-dotenv
3pipenv install psycopg2-binary

Setup Postgres and RabbitMQ

Create a docker-compose.yaml to run Postgres and Rabbitmq in the background.

1version: '3'
3 db:
4 image: postgres:13
5 env_file:
6 - .env
7 ports:
8 - 5432:5432
9 rabbitmq:
10 image: rabbitmq
11 ports:
12 - 5672:5672


  • To include the app in our project, we need to add a reference to its configuration class in the INSTALLED_APPS setting in newsletter_site/
2 'newsletter.apps.NewsletterConfig',
3 'django.contrib.admin',
4 'django.contrib.auth',
5 'django.contrib.contenttypes',
6 'django.contrib.sessions',
7 'django.contrib.messages',
8 'django.contrib.staticfiles',
  • We need to tell Celery how to find RabbitMQ. So, open and add this line:


  • We need to configure database settings:
1uri = os.getenv('DATABASE_URL')
3result = urlparse(uri)
5database = result.path[1:]
6user = result.username
7password = result.password
8host = result.hostname
9port = result.port
12 'default': {
13 'ENGINE': 'django.db.backends.postgresql',
14 'NAME': database,
15 'USER': user,
16 'PASSWORD': password,
17 'HOST': host,
18 'PORT': port,
19 }
  • We need to configure the SMTP server in . SMTP server is the mail server responsible to deliver emails to the users. For development, you may use a Gmail SMTP server, but this has limits and will not work if you have 2 FA. You can refer to this article. For production, you can use commercial services such as sendgrid.
1EMAIL_BACKEND = 'django.core.mail.backends.smtp.EmailBackend'
2EMAIL_HOST = os.getenv('EMAIL_HOST')
3EMAIL_USE_TLS = bool(os.getenv('EMAIL_USE_TLS'))
4EMAIL_PORT = os.getenv('EMAIL_PORT')

For your reference, you can see the here.

Create env file

Create a .env file and assign the secrets.



We need to set up Celery with some config options. Create a new file called inside newseletter_site directory:

1import os
3from celery import Celery
5os.environ.setdefault('DJANGO_SETTINGS_MODULE', 'newsletter_site.settings')
7app = Celery('newsletter_site')
9app.config_from_object('django.conf:settings', namespace='CELERY')

Design and Implement Models & Configure Admin

This is the schema we are trying to build. The schema is implemented here. Create a newsletter/ with the same content.

Model Design

We need a UI to manage the newsletter. We will be using Django Admin for this purpose. Create a newsletter/ with the contents of this file.

Register URL for admin in newsletter_site/

1urlpatterns = [
2 path('admin/',,

Run the app

Run docker-compose to start the dependencies:

docker-compose up

Generate migrations for our models:

python makemigrations

To apply generated migrations to database run:

python migrate

To create a user for login run the following command and provide your details:

python createsuperuser

Run the following command to run the app and open to open Django Admin :

python runserver

Django Admin

Run celery:

celery -A newsletter_site worker --loglevel=INFO Add a newsletter and a subscriber and subscribe them to it. Create an issue and send it. If everything is fine you will see an issue arriving in your email.

How does it work?

When we click send the following action gets executed:

1def send(modeladmin, request, queryset):
2 for issue in queryset:
3 tasks.send_issue.delay(
6send.short_description = "send"

This code is responsible to queue up a new task to send an issue using celery. It publishes the task to RabbitMQ.

2def send_issue(issue_id):
3 issue = Issue.objects.get(pk=issue_id)
4 for subscription in Subscription.objects.filter(newsletter=issue.newsletter):
5 send_email.delay(, issue.title, issue.content)
8def send_email(email, title, content):
9 send_mail(
10 title,
11 content,
12 '',
13 [email],
14 fail_silently=False,
15 )

The Celery worker uses these tasks. When the producer produces a task the worker runs the corresponding task.

When we publish the send_issue task we determine the subscriber for the newsletter and publish sub-tasks to send the actual email. This strategy is called fan-out. Fan out is useful as it allows us to retry sending emails to a single user in case of a failure.


In this post, we saw how to use RabbitMQ as a message queue with Celery and Django to send bulk emails. This is a good fit where message queues are appropriate. Use message queue if the request is indeterministic or the process is long-running and resource-intensive. You can find the finished project here

Originally Published here.

More articles from Nancy Chauhan

Top 10 Productive Hacks for Software Developers

Recently, I posted a tweet asking all the amazing software developers in my network to tell about the hacks they use to keep themselves productive. I have compiled the wonderful and helpful solutions I received. Hope you find them useful as well.

April 17th, 2021 · 4 min read

2020 at a glance

This year was a rollercoaster for me. But guess what? I am retrospecting 2020, and I think I did a good job.

December 31st, 2020 · 4 min read
© 2018–2022 Nancy Chauhan
Link to $ to $ to $ to $ to $