DjangoUnboxed
Production ready django based starter kit application.
With collaborative efforts from the CTO office, we have prepared this Django Boilerplate. This boilerplate is aimed to be generic in a way that it solves most of the common problems that are faced in development process, specific to Python-Django development. Moreover it also has the best practices followed in web app development catering the microservice architecture.
Technology
- Django 3
- Django Rest Framework
- Fully Dockerised Setup (configuration driven)
- Smart Logging (DB, Requests)
- Elastic Solutions (FileBeat, MetricBeat and HeartBeat)
- Celery
- Rabbit MQ
- JWT
- Unit tests
- Swagger
Description
- Django :- We have opted for Django, as we wanted to have a strong framework and complete ORM solution since we are planning to go with Python and with SQL based database
- Django Rest Framework :- We opted for DRF as this is the best solution available for REST APIs for Python.
- Fully Dockerised Setup :- Docker has become an integral part of the setup these days. All the settings are yml driven and are configurable as per the business need. One can easily turn off/on configurations/services via this.
- Smart Logging and EK :- We have integrated customized logger interface, tracking DB Logs, Request Logs and App Logs with our own hand tailored logger implementation that enables us to predict from the logs if the application is going to face issues/problems. Moreover we have integrated the EK stack in this that enables us to harness great analytical and searching tool Kibana. Basically the boilerplate is designed to transfer all the logs on Kibana, this is happening via Filebeat. All the logs can then be queried and analytics can be drawn from them.
- System Monitoring :- We have included integration of MetricBeat and HeartBeat, they are efficient solutions to monitor system levels and vitals.
- Celery :- The most preferred delayed task runner when handling asynchronous tasks with django applications. Very robust & easy to integrate.
- Rabbit MQ :- When working with microservices, the quintessential requirement is of inter service communication, this where message broker comes into picture. Since we need to avoid the synchronous dependencies of the REST communication. Hence we have provisioned dedicated pub/sub configurations that constantly monitor the Rabbit MQ and execute tasks in asynchronous way.
- JWT :- The reason why we selected JWT is that, it is used is to prove that the sent data was actually created by an authentic source and this is the most widely used mechanism while communication is happening over REST APIs.
- Unit Test :- We have included the unit tests in such a way that the APIs are stubbed and there is no incorporation of Databases, thus avoiding the overheads related with Databases. Since we aim to have pure unit tests hence we have provisioned stubbed methods for unit tests.
- Swagger :- Though for REST APIs we can get a good document available via Django REST Framework, but that is limited in some ways like publishing. We selected swagger, so that APIs can be published and be tested externally.
Application Structure
|____boilerplate
| |____config
|____boilerplate_app
| |____migrations
|____compose
|____publisher_subscriber
|____requirements
Running the server locally
- Clone this repo
- Install python3.10
- Intall dependencies:
pip install -r requirements.txt
- Run the server:
python manage.py runserver
Docker Setup
- Install docker compose
- Run docker:
Create master_app.log file inside root directory
docker-compose build
docker-compose up
- To check the server, open
http://localhost:8000/
Contributors
Rishabh Shah
License
This project is licensed under the terms of the MIT license.