Mastering Production-Grade Docker Compose Your Essential Guide

To ensure isolation and persistent data, custom networks, volumes, and .env files are crucial.

1. Define a custom network: This isolates your services.

networks:
  my_app_net:
    driver: bridge

2. Define a named volume: This stores data persistently.

volumes:
  my_app_data:

3. Create an .env file: Put sensitive info here, e.g., DB_PASSWORD=secret.

# .env file
DB_USER=admin
DB_PASSWORD=supersecret

4. Integrate into docker-compose.yml:

version: '3.8'
services:
  web:
    image: nginx:latest
    volumes:
      - my_app_data:/var/www/html # Bind mount an anonymous volume
    environment:
      - DB_USER=${DB_USER}
      - DB_PASSWORD=${DB_PASSWORD}
    networks:
      - my_app_net

networks:
  my_app_net:
    driver: bridge

volumes:
  my_app_data:

5. Run Compose:

docker-compose up -d

Verification:

• Check networks:

  docker network ls

  You should see my_app_net.

• Check volumes:

  docker volume ls

  You should see <project_name>_my_app_data.

• Check service environment variables:

  docker exec <container_id> env

Common Failure: Missing .env file or wrong variable name.
Fix: Ensure .env exists and variable names match those in docker-compose.yml exactly.

Health checks ensure your services are truly alive. Restart policies dictate how Docker Compose reacts to service failures.

To add a health check:

version: '3.8'
services:
  web:
    image: nginx
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 5s

The test command evaluates the service’s health. CMD runs a command, NONE disables it.

To define restart policies:

services:
  web:
    image: nginx
    restart: always

Common policies:

• no: Never restart.
• on-failure: Restart only if the container exits with a non-zero code.
• always: Always restart, even on graceful shutdown.
• unless-stopped: Always restart unless explicitly stopped.

Verify health status with: docker-compose ps or docker ps.

Failure mode: Health checks too strict or too lenient. Adjust interval, timeout, retries, and start_period to match application startup and response times. Incorrect restart policies can lead to services not recovering or constantly restarting. Choose a policy that matches expected service behavior.

These features, combined with robust networking, volumes, and environment management, create a resilient, self-healing Docker Compose environment.