1mr-newton / gunicorn-flask-docker

A small and fast image with Gunicorn to deploy your Flask applications!

Gunicorn Server with Flask in Docker

This is a small and fast Docker image to fast deploy your Flask applications via Gunicorn server.

All images are based on Python-Alpine images.

GitHub Repository Page: https://github.com/iamelin/gunicorn-flask-docker

Docker Hub Page: https://hub.docker.com/r/amelin/gunicorn-flask

Supported tags

• python3.9
• python3.8
• python3.7
• python3.6

Quick Start

Deploying your application is very simple! Just do the following:

1. Create a directory app with your Python application. The main file must be named main.py and the Flask object must be named app.
2. Pull the image from Docker Hub: docker pull amelin/gunicorn-flask:python3.8
3. Create Dockerfile with the following content:

FROM amelin/gunicorn-flask:python3.8

ADD ./app/ /app

That's all! Now build your image and start a new container. If the file /app/requirements.txt exists, packages will be installed automatically before starting the application. Then you can send requests to your application that listens port 5000.

Configuration

You can configure Gunicorn and image using the following environment variables:

• PORT — port that the server listens (default value: 5000)
• WORKERS — number of workers that handle requests (default value: number of cores of your processor * 2 + 1)
• LOGLEVEL — level of logs (default value: info)
• ACCESS_LOG — file which the server writes access logs to (default no access logs)
• ERROR_LOG — file which the server writes error logs to (default value: stdout)
• MODULE_NAME — a Python module that contains a Flask object (default value: main, file main.py in the working directory /app)
• APP_NAME — a Flask object which handles requests (default value: app)
• GUNICORN_CONFIG — a Python script that contains the configuration of a Gunicorn instance (default value: gunicorn.conf.py in the working directory /app)
• TZ — timezone (default value: UTC)

Also the application directory /app contains the file prestart.sh that runs before starting a container. You can use it to prepare your application to work (e.g. run migrations etc).