Containerizing Full-Stack Systems with Docker and Nginx

Deploying multi-tier applications across microservices or hybrid static/dynamic stacks often suffers from over-engineering. In many environments, a simple, declarative Docker Compose setup coupled with Nginx provides superior reliability and observability compared to heavyweight Kubernetes clusters.

The Architecture Pattern

In this architecture, Nginx acts as the single ingress gateway for our domain:

[ Incoming Requests (Port 443 / 80) ]
                  |
                  v
       [ Nginx Reverse Proxy ]
        /         |          \
       v          v           v
  [ Static/PHP ] [ Astro Blog ] [ Backend Services ]
   (Port 9000)    (Port 4321)      (DB / Queues)

Nginx Subpath Routing

To route traffic to our Astro blog running in its own container, we configure Nginx to proxy /blog requests directly to the blog service name:

# docker/nginx/default.conf
location /blog {
    proxy_pass http://blog:4321;
    proxy_http_version 1.1;
    proxy_set_header Upgrade $http_upgrade;
    proxy_set_header Connection "upgrade";
    proxy_set_header Host $host;
    proxy_set_header X-Real-IP $remote_addr;
    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    proxy_set_header X-Forwarded-Proto $scheme;
}

The Docker Compose Definition

The blog service definition is straightforward:

services:
  blog:
    container_name: codemonkeyg-blog
    restart: unless-stopped
    build:
      context: .
      dockerfile: docker/blog/Dockerfile
    volumes:
      - ./blog:/app
      - /app/node_modules
    ports:
      - "4321:4321"
    environment:
      - NODE_ENV=development

Using volume mounts allows local changes to Markdown files or Astro components to trigger instant Hot Module Replacement (HMR) without restarting the container.

Key Benefits

• Isolation: The blog’s Node.js runtime is completely decoupled from PHP-FPM or static assets.
• Portability: Any developer can clone the repository and run docker compose up -d to get the entire website and blog running locally.
• Zero-downtime upgrades: Updating blog dependencies or node versions never affects the primary application.