Skip to content

deploy.en.md

Latest commit

 

History

History
165 lines (119 loc) · 3.99 KB

deploy.en.md

File metadata and controls

165 lines (119 loc) · 3.99 KB

Deployment & Configuration

This page contains technical details for running DataLab Web, including Python/Flask setup and environment variables.

Stack

  • Backend: Flask (Python web framework)
  • Numerics: mpmath (multi-precision floating point)
  • Frontend: plain HTML + CSS + JavaScript
  • LaTeX engines: pdflatex / xelatex

Run Locally

Requirements

  • Python 3.8+
  • Python dependencies (see web_requirements.txt)

Steps

# 1. Enter the project directory
cd /path/to/data_extrapolation_source

# 2. Install dependencies
pip install -r web_requirements.txt

# 3. Start the server
python app_web/server.py

Open in Browser

Visit: http://127.0.0.1:8000

Environment Variables (optional)

# Bind address (default 127.0.0.1)
export DATALAB_HOST=0.0.0.0

# Port (default 8000)
export DATALAB_PORT=5000

# Enable debug (development only!)
export DATALAB_DEBUG=1

Production Deployment

Option 1: Gunicorn + Nginx

Note: Gunicorn is POSIX-only (Linux/macOS). It will not run on native Windows because it depends on fcntl.

# 1. Install gunicorn
pip install gunicorn

# 2. Start gunicorn (4 workers)
gunicorn -w 4 -b 127.0.0.1:8000 app_web.server:app

# 3. Configure nginx reverse proxy
# /etc/nginx/sites-available/datalab
server {
    listen 80;
    server_name your-domain.com;

    location / {
        proxy_pass http://127.0.0.1:8000;
        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;
    }
}

# 4. Restart nginx
sudo systemctl restart nginx

Option 2: systemd Service

Create /etc/systemd/system/datalab-web.service:

[Unit]
Description=DataLab Web Application
After=network.target

[Service]
Type=notify
User=www-data
WorkingDirectory=/path/to/data_extrapolation_source
Environment="DATALAB_WEB_SECRET=your-secret-key-here"
Environment="DATALAB_HOST=127.0.0.1"
Environment="DATALAB_PORT=8000"
ExecStart=/usr/bin/gunicorn -w 4 -b 127.0.0.1:8000 app_web.server:app
Restart=always

[Install]
WantedBy=multi-user.target

Enable and start:

sudo systemctl enable datalab-web
sudo systemctl start datalab-web

Windows (native): Waitress

Gunicorn does not support Windows. Use Waitress instead:

pip install waitress
waitress-serve --listen=192.168.85.1:8000 --threads=8 app_web.server:app

If you need multi-process workers on Windows, consider running the service in WSL2/Docker and using Gunicorn there.

Built-in Docs (/docs)

The web app provides built-in docs via /docs and /docs/<page>:

  • Documentation files are loaded only from docs/web/ with a strict whitelist (prevents path traversal)
  • File naming: docs/web/<page>.<lang>.md where lang ∈ {zh,en}
  • Language selection priority:
    1. Query parameter ?lang=zh or ?lang=en
    2. Cookie datalab_lang
    3. Default zh
  • If an English page is missing, an English placeholder page is shown (no fallback to Chinese)

Configuration

Required (production)

DATALAB_WEB_SECRET — Flask secret key

  • Used for session signing and CSRF tokens
  • Must be random and kept secret (at least 32 characters)
  • Default datalab-web-dev is for development only

Generate a key:

python -c "import secrets; print(secrets.token_hex(32))"

Optional

Variable Default Description
DATALAB_HOST 127.0.0.1 Bind address (0.0.0.0 = all interfaces)
DATALAB_PORT 8000 Port
PORT 8000 Fallback port (cloud platforms)
DATALAB_DEBUG 0 Debug mode (1 enables, production must be 0)

Security Notes

  • Do not hard-code secrets into source code
  • Use HTTPS in production
  • Limit upload sizes (built-in validation exists)
  • Keep LaTeX compilation sandboxed and time-limited

Logs & Debugging

  • Run with DATALAB_DEBUG=1 in development only
  • For systemd deployments, use journalctl -u datalab-web -f