URA-Backend

Backend for image classification tool using Django, Celery and RabbitMQ.

Email huhenry50@gmail.com for any questions.

To interact with the server, make sure the frontend is running as well.

Table of Contents

• Local Usage
  • With Docker
  • Without Docker
• Production Environment
• TODO

Local Usage

The server can be run using Docker and without Docker.

With Docker

The images are built locally. For the curious, details are in Dockerfile and docker-compose.yml.

• 1. First Time Setup

  From a new terminal, in your local cloned repository directory, execute:

  $ docker-compose build

• 2. Run Server

  $ docker-compose up -d
  $ docker-compose run web manage.py makemigrations
  $ docker-compose run web manage.py migrate

• 3. Stop Server

  $ docker-compose down

Without Docker

If you do not wish to use the Docker image, this section provides instructions on how to run the server manually.

Note: this section still uses Docker for RabbitMQ. You can follow the below RabbitMQ setup or set up RabbitMQ locally without a RabbitMQ image.

Environment Variables

In your local repo, create a .env file and place these contents inside:

PRODUCTION=False
DEBUG=True

DB_ENGINE=django.db.backends.sqlite3
DB_NAME={place your desired sqlite file path here}

# Frontend host and port
CORS_ALLOWED_ORIGINS=http://localhost:3000
 
# OPTIONAL email configuration, check the SMTP servers and ports 
EMAIL_HOST={email server host here, ex. smtp.gmail.com} 
EMAIL_PORT={email server port here, ex. 587}
EMAIL_HOST_USER={your email here}
EMAIL_HOST_PASSWORD={your email password here}

# RabbitMQ host and port
CELERY_BROKER_URL=amqp://guest:guest@localhost:5672//

BE CAREFUL: DO NOT SHARE THIS .env FILE.

First Time Setup

This setup uses pipenv as the Python virtual environment.

From a new terminal, in your local cloned repository directory, execute:

$ pipenv shell
$ pipenv install
$ cd uClassify/
$ python manage.py makemigrations
$ python manage.py migrate

Run Server

Continuing from above, execute:

$ python manage.py runserver

For every fresh run, execute:

$ pipenv shell
$ cd uClassify/
$ python manage.py runserver

RabbitMQ

By default, the RabbitMQ image is used by celery. You can pull any RabbitMQ image you like from Docker Hub.

ex.

$ docker pull rabbitmq

Then run RabbitMQ (either through the Docker desktop app or from your command line).

IMPORTANT: make sure the CELERY_BROKER_URL is pointing to the RabbitMQ port, by default it is 5672

Celery Worker

From a new terminal, in your local cloned repository directory, execute:

$ pipenv shell
$ cd uClassify
$ celery -A uClassify worker -l INFO -P solo

Production Environment

The EC2 instance is currently down.

AWS powers the server hosted on my domain, utilizing EC2, RDS and S3.

TODO

• code cleanup
• improve error handling
• (prod) provide classification model downloads