Skip to main content

Containerizing Python applications with Docker (Part 1): Basics

Packaging clarified how to publish artifacts; Docker answers how to ship reproducible environments: OS packages, interpreters, pinned dependencies, plus your code share one immutable image digest.

Install Docker Desktop (Windows/macOS) or Docker Engine on Linux before following along.

📚 Prerequisites​

  • Working Dockerfile vocabulary from DevOps teammates helps but is not mandatory.

🎯 What you'll learn​

  • Author a slim Python image leveraging layer caching.
  • Pass runtime configuration purely via env vars/secrets—not baked secrets.

Minimal Python service image​

FROM python:3.12-slim

ENV PYTHONDONTWRITEBYTECODE=1 \
    PYTHONUNBUFFERED=1

WORKDIR /app

COPY pyproject.toml README.md ./
COPY src ./src

RUN pip install --no-cache-dir .

CMD ["python", "-m", "myservice.worker"]

slim images reduce attack surface versus full python installs still large enough for many APIs.

Build & run locally​

docker build -t myservice:local .
docker run --rm -e LOG_LEVEL=debug myservice:local

💡 Key takeaways​

  • Order COPY lines from least-changing to most-changing so dependency layers stay cached.
  • Never bake .env files with prod secrets—mount or inject via orchestrators.

➡️ Next steps​

Specialize Flask layouts in **Docker Part 2: Flask Dockerfile**.