containerization-and-deploymentlisted

# Containerization and Deployment Docker containerization, multi-role dispatch, secrets handling, and local dev infrastructure patterns from production services. ## When to use - Containerizing FastAPI/Python backend services - Implementing one-image-many-roles pattern (server/consumer/worker/cron from same Dockerfile) - Writing certs/keytabs from environment variables for Kafka SASL_SSL/Kerberos - Setting up local dev with docker-compose (postgres/redis/kafka/temporal) - Deploying to Cloud Run, GKE, or other container platforms - Handling secrets and credentials in containerized environments - Configuring health checks and readiness probes ## Core conventions ### Multi-Stage Dockerfile Pattern **Builder + runtime separation**: Use a builder stage to install build dependencies and compile wheels; copy only the virtual environment and application code to a minimal runtime stage. **Alpine for minimal footprint**: Use `python:3.10-alpine` or `python:3.11-slim` base images; install only runtime dependencies in the final stage. Alpine offers minimal footprint; slim provides broader compatibility. **Dependency caching**: Copy `requirements.txt` before application code; leverage Docker layer caching to avoid reinstalling dependencies on every code change. **Security hardening**: Remove `.git` directory after capturing commit SHA; use `--no-cache-dir` with pip; set `PYTHONUNBUFFERED=1` for immediate log output. ### Entrypoint MODE Dispatch **Single entrypoint, multiple rol