Optimizing Docker builds for Python applications

Speaker

Dmitry Figol

Material

• video
• slide

Note

• Docker Concept

• Optimize Goal
  1. Reducing image size
  2. Reducing build time
• Priority
  • Fast build during development → Use slim-stretch
  • Small size during production → Use alpine

build deps are only needed when compiling but not runtimeImprovement

• Use specific COPY statement instead of COPY ..
• .dockerignore

delete build deps doesn't shrink image size

• Docker Layer
  • Instructions create read-only layers
  • A new layer can't be smaller than the previous layer
  • Layers are cached and can be re-used for subsequent builds
  • Layers introduce some overhead
• Tips
  • Combine multiple RUN statements into one
  • delete files in the same layer (instruction) where they were added
  • arrangement statement from the least changing to the most changing (system-level -> tools -> Python deps -> source code)
  • don't save deps to cache
    • pip: --no-cache-dir
    • apk: --no-cache

complicate Dockerfile and no cache during build

• Docker multi-stage
  • Build an intermediate image with all build deps and install your app
  • Copy the result to a fresh image

Other Tips

• Bind mount source code instead of COPY in local dev env
• PYTHONUNBUFFERED=1 -> print to stdout without buffering
• PYTHONDONTWRITEBYTECODE=1 -> no *.pyc files

Share on: Twitter ❄ Facebook ❄ Email

---