Skip to content

Deploy Flask with Docker on Elastic Beanstalk

This tutorial shows the Docker deployment path for a Flask application on Elastic Beanstalk. It compares native Python platform usage with single-container Docker options.

Prerequisites

  • Docker installed locally.
  • Existing Flask app with application.py.
  • Elastic Beanstalk application ready for Docker platform branch.

What You'll Build

You will build:

  • A Docker image for Flask + Gunicorn.
  • Dockerrun.aws.json metadata for single-container deployment.
  • Decision criteria for Docker versus native Python platform.

Steps

  1. Create Dockerfile for Flask runtime.
FROM public.ecr.aws/docker/library/python:3.11-slim

WORKDIR /app
COPY requirements.txt requirements.txt
RUN pip install --no-cache-dir -r requirements.txt
COPY . .
EXPOSE 8000
CMD ["gunicorn", "--bind", ":8000", "--workers", "2", "application:application"]
  1. Build local image for validation.
docker build --tag flask-eb:latest .
docker run --rm --publish 8000:8000 flask-eb:latest
  1. Add Dockerrun.aws.json for Elastic Beanstalk single container deployment.
{
    "AWSEBDockerrunVersion": "1",
    "Image": {
        "Name": "<account-id>.dkr.ecr.ap-northeast-2.amazonaws.com/flask-eb:latest",
        "Update": "true"
    },
    "Ports": [
        {
            "ContainerPort": "8000"
        }
    ]
}
  1. Initialize Elastic Beanstalk for Docker platform if creating separate environment.
eb init --platform "Docker running on 64bit Amazon Linux 2023" --region "$REGION"
  1. Deploy Docker bundle.
eb deploy --staged

  1. Choose runtime mode based on AWS-documented tradeoffs:

  2. Native Python platform for direct language-platform features.

  3. Docker when you need container-level dependency and runtime control.
flowchart TD
    A[Flask Source] --> B[Native Python Platform]
    A --> C[Dockerfile Build]
    C --> D[Container Image]
    D --> E[Dockerrun.aws.json]
    E --> F[Elastic Beanstalk Docker Environment]

Verification

Validate container deployment behavior:

eb status "$ENV_NAME"
eb logs --all
curl --verbose "http://$CNAME"

Expected outcomes:

  • Environment health becomes green after deployment.
  • Logs show container startup and Gunicorn bind on port 8000.
  • Root endpoint responds from containerized Flask app.

See Also

Sources