Scaling — From Laptop to Cloud

Pytigon applications scale horizontally — from a single-developer SQLite setup to a multi-server PostgreSQL cluster behind a load balancer, without changing application code.

Scaling Dimensions

Vertical Scaling (Bigger Server)

The simplest path: give Pytigon more CPU, RAM, and faster disks. Pytigon benefits from:

  • More RAM — Larger QuerySet caches, more concurrent template renders
  • Faster CPU — Quicker .ihtml → .html compilation
  • SSD storage — Faster virtual filesystem operations

Horizontal Scaling (More Servers)

Pytigon's stateless design (all state in the database and VFS) enables horizontal scaling:

                 [Load Balancer]
                  /     |     \
            [App 1] [App 2] [App 3]
                  \     |     /
               [PostgreSQL] [Redis]

Production Architecture

Internet → Nginx (SSL termination, static files)
              ├─→ Gunicorn (WSGI: HTTP requests)
              └─→ Daphne (ASGI: WebSockets)
                     ↓
                  Django
                     ↓
              PostgreSQL + Redis

Nginx Configuration

upstream pytigon_app {
    server 127.0.0.1:8000;
    server 127.0.0.1:8001;
    server 127.0.0.1:8002;
}

server {
    listen 443 ssl;
    server_name myapp.example.com;

    location /static/ {
        alias /var/www/pytigon/static/;
        expires 30d;
    }

    location / {
        proxy_pass http://pytigon_app;
        proxy_set_header Host $host;
        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    }

    location /ws/ {
        proxy_pass http://pytigon_ws;
        proxy_http_version 1.1;
        proxy_set_header Upgrade $http_upgrade;
        proxy_set_header Connection "upgrade";
    }
}

Database Scaling

  • SQLite — Single-server only. Good for development, small deployments.
  • PostgreSQL — Multi-server. Use connection pooling (PgBouncer), read replicas for reporting queries.
  • Connection pooling — Django's CONN_MAX_AGE setting keeps database connections alive between requests.

Caching

Cache Type Backend Use For
Template cache File/Memcached Compiled .ihtml → .html
Query cache Redis Frequently accessed data
Session cache Redis User sessions (required for multi-server)
Static files Nginx/CDN CSS, JS, images

Background Tasks at Scale

For multi-server deployments, configure a dedicated task worker:

# settings_app.py
TASK_BACKEND = 'redis'  # or 'rabbitmq'
TASK_WORKER_COUNT = 4

Tasks are distributed across workers using Redis as the message broker.

?: Scaling ?: Scaling