django-deploy-toolkit 0.2.0

Auto-generate and install Gunicorn, Nginx, Celery & Celery Beat configuration for Django projects on Ubuntu/Debian.

django-deploy-toolkit

Auto-generate and install Gunicorn, Nginx, Celery & Celery Beat configuration for Django projects on Ubuntu/Debian.

django-deploy-toolkit detects your Django project's configuration, generates production-ready systemd service/socket files and Nginx server blocks, and installs them — all with a single command. It handles edge cases gracefully, supports dry-run mode, and includes full rollback on failure.

---

Created By

Antu Saha — creator and lead maintainer of this package.

We also appreciate all the open-source contributors who help improve django-deploy-toolkit. Contributions of any kind are welcome — bug reports, feature requests, documentation improvements, and code contributions. See the Contributing section below to get started.

---

Requirements

• OS: Ubuntu / Debian Linux
• Python: 3.8 or higher
• Django project with a manage.py file
• Gunicorn installed in the project's virtual environment
• Nginx installed on the server (sudo apt install nginx)
• systemd available (standard on Ubuntu/Debian)

---

Installation

pip install django-deploy-toolkit

Or install from source:

git clone https://github.com/antusaha970/django-deploy-toolkit.git
cd django-deploy-toolkit
pip install -e .

---

Quick Start

# Navigate to your Django project root (where manage.py lives)
cd /path/to/your/django/project

# Activate your project's virtual environment
source venv/bin/activate

# Run the setup
django-deploy setup

That's it. The tool will:

1. Auto-detect your project configuration
2. Show you what it found and ask for confirmation
3. Generate and install the config files
4. Enable and start the services
5. Test and reload Nginx
6. Detect Celery — if Celery is installed, it hints you to run django-deploy celery-setup

---

What Gets Auto-Detected

Setting	Detection Method
Project Name	Basename of the current directory (sanitized)
Project Path	Current working directory (searches for manage.py)
WSGI Module	Parsed from manage.py → DJANGO_SETTINGS_MODULE, or scans for wsgi.py
User	$USER env var → $LOGNAME → pwd.getpwuid()
Group	Primary group of the detected user via grp.getgrgid()
Python Path	sys.executable (virtualenv-aware)
Server IP	/etc/hosts → socket.gethostbyname() → UDP socket trick → _ (catch-all)
Workers	(2 × CPU cores) + 1, capped at 9 (via psutil)
Static Root	Parsed from python manage.py diffsettings output
Media Root	Same as Static Root

---

Celery & Celery Beat Support

django-deploy-toolkit can also generate and install systemd service files for Celery worker and Celery Beat — keeping your background task infrastructure managed by systemd alongside Gunicorn.

What Gets Detected (Celery)

Check	Detection Method
Celery installed	python -c "import celery" using the project's Python
Celery app module	Scans for celery.py in project packages, or Celery(...) instantiation
Celery Beat enabled	CELERY_BEAT_SCHEDULE in settings, or django_celery_beat in INSTALLED_APPS
Broker URL	CELERY_BROKER_URL / BROKER_URL from Django settings
Redis installed	redis-server or redis-cli on $PATH

Celery Quick Start

# After running django-deploy setup (or standalone):
django-deploy celery-setup

The tool will:

1. Auto-detect your Celery configuration
2. Show you a summary and ask for confirmation
3. Generate systemd service files for Celery worker (and Beat if detected)
4. Install, enable, and start the services
5. Remind you to adjust the worker concurrency (defaults to 1)

Celery CLI Options

django-deploy celery-setup [OPTIONS]

Options:
  --dry-run              Show what would be done without doing it
  --verbose / --no-verbose  Show or hide detailed output
  --no-confirm           Skip the confirmation prompt
  --help                 Show this message and exit

Edge Cases Handled

• Redis not installed locally — If your broker URL uses redis:// but redis-server isn't found, a warning is shown. This is non-blocking because Redis may run on a remote host.
• Celery without Beat — Only the worker service is generated.
• Beat with django_celery_beat — The --scheduler django_celery_beat.schedulers:DatabaseScheduler flag is automatically added to the Beat service.
• No Celery found — A helpful message is shown and the command exits cleanly.
• Celery app module not found — The user is prompted to enter it manually.

---

Dry-Run Mode

Preview everything that would happen without making any changes:

django-deploy setup --dry-run
django-deploy celery-setup --dry-run

Example output:

🚀 django-deploy-toolkit — Setup

  [DRY RUN] Would write: /etc/systemd/system/myproject.socket
  [DRY RUN] Would write: /etc/systemd/system/myproject.service
  [DRY RUN] Would write: /etc/nginx/sites-available/myproject
  [DRY RUN] Would remove: /etc/nginx/sites-enabled/default
  [DRY RUN] Would remove: /etc/nginx/sites-available/default
  [DRY RUN] Would create symlink: /etc/nginx/sites-enabled/myproject -> /etc/nginx/sites-available/myproject
  [DRY RUN] Would run: systemctl daemon-reload
  [DRY RUN] Would run: systemctl enable myproject.socket
  [DRY RUN] Would run: systemctl start myproject.socket
  [DRY RUN] Would run: systemctl enable myproject.service
  [DRY RUN] Would run: nginx -t
  [DRY RUN] Would run: systemctl reload nginx

---

Manual Overrides

All CLI flags:

django-deploy setup [OPTIONS]

Options:
  --dry-run              Show what would be done without doing it
  --project-path PATH    Override the detected project path
  --project-name NAME    Override the detected project name
  --no-confirm           Skip the confirmation prompt (for scripts/CI)
  --verbose / --no-verbose  Show or hide detailed output
  --version              Show version and exit
  --help                 Show this message and exit

Other commands:

django-deploy detect             # Just run detection, no install
django-deploy generate           # Generate files to current directory
django-deploy generate --output-dir ./configs  # Generate to a specific directory
django-deploy rollback --name myproject        # Rollback a previous install
django-deploy celery-setup       # Generate and install Celery systemd services
django-deploy celery-setup --dry-run  # Preview Celery setup

---

Generated File Examples

Gunicorn Socket (/etc/systemd/system/myproject.socket)

[Unit]
Description=gunicorn socket for myproject

[Socket]
ListenStream=/run/myproject.sock

[Install]
WantedBy=sockets.target

Gunicorn Service (/etc/systemd/system/myproject.service)

[Unit]
Description=gunicorn daemon for myproject
Requires=myproject.socket
After=network.target

[Service]
User=deploy
Group=deploy
WorkingDirectory=/home/deploy/myproject
ExecStart=/home/deploy/myproject/venv/bin/python -m gunicorn \
          --access-logfile - \
          --workers 5 \
          --bind unix:/run/myproject.sock \
          myproject.wsgi:application

[Install]
WantedBy=multi-user.target

Nginx Config (/etc/nginx/sites-available/myproject)

server {
    listen 80;
    server_name 192.168.1.100;

    location /static/ {
        root /home/deploy/myproject;
    }

    location /media/ {
        root /home/deploy/myproject;
    }

    location / {
        include proxy_params;
        proxy_pass http://unix:/run/myproject.sock;
    }
}

Celery Worker (/etc/systemd/system/myproject-celery.service)

[Unit]
Description=Celery Worker for myproject
After=network.target

[Service]
Type=forking
User=deploy
Group=deploy
WorkingDirectory=/home/deploy/myproject
ExecStart=/home/deploy/myproject/venv/bin/python -m celery -A myproject.celery worker \
          --loglevel=info \
          --concurrency=1 \
          --pidfile=/run/celery/myproject-worker.pid \
          --logfile=/var/log/celery/myproject-worker.log
ExecStop=/bin/kill -s TERM $MAINPID
RuntimeDirectory=celery
Restart=always

[Install]
WantedBy=multi-user.target

Celery Beat (/etc/systemd/system/myproject-celerybeat.service)

[Unit]
Description=Celery Beat Scheduler for myproject
After=network.target

[Service]
Type=simple
User=deploy
Group=deploy
WorkingDirectory=/home/deploy/myproject
ExecStart=/home/deploy/myproject/venv/bin/python -m celery -A myproject.celery beat \
          --loglevel=info \
          --pidfile=/run/celery/myproject-beat.pid \
          --logfile=/var/log/celery/myproject-beat.log \
          --scheduler django_celery_beat.schedulers:DatabaseScheduler
ExecStop=/bin/kill -s TERM $MAINPID
RuntimeDirectory=celery
Restart=always

[Install]
WantedBy=multi-user.target

---

Default Nginx Config Removal

During installation, django-deploy-toolkit removes the default Nginx configuration files:

• /etc/nginx/sites-enabled/default
• /etc/nginx/sites-available/default

Why? The default Nginx config listens on port 80 and can conflict with your project's server block. Removing it ensures your Django project is the one handling requests.

Safety measures:

• Both files are backed up to a temporary directory before deletion
• The tool warns you and pauses for 3 seconds before proceeding (press Ctrl+C to cancel)
• If you skip if they don't exist — no error
• During rollback, the backed-up files are restored to their original locations

---

Rollback

If any installation step fails, django-deploy-toolkit automatically rolls back all changes:

1. Files created during the run are deleted
2. Symlinks created during the run are removed
3. The default Nginx config is restored from backup (if it was removed)
4. Services that were started are stopped and disabled
5. Systemd daemon is reloaded
6. Nginx config is tested and reloaded

You can also manually rollback a previous installation:

django-deploy rollback --name myproject

---

Contributing

Contributions are welcome! Whether it's a bug fix, new feature, or documentation improvement — all help is appreciated.

# Clone the repository
git clone https://github.com/antusaha970/django-deploy-toolkit.git
cd django-deploy-toolkit

# Install dev dependencies
make install-dev

# Run tests
make test

# Run tests with coverage
make coverage

---

Credits

• Antu Saha — Creator and lead maintainer
• All open-source contributors — Thank you for helping make this project better! 🙏

---

License

MIT — see LICENSE for details.