如何在MEAN栈Node.js应用中部署运行Python脚本及安装依赖

Nice question—getting Python dependencies sorted when deploying a MEAN stack app that calls Python scripts is a super common hurdle, but there are reliable ways to handle it based on where you’re hosting. Let’s walk through the most practical approaches:

1. Traditional VPS/Cloud Servers (e.g., AWS EC2, DigitalOcean Droplet) #

If you’re deploying to a raw server, you’ll need to manually set up the Python environment and install dependencies:

• First, make sure Python and pip are installed on the server (most Linux distros have them pre-installed, but updating is a good idea):

  # For Debian/Ubuntu-based servers
  sudo apt update && sudo apt install python3 python3-pip -y
  
  # For RHEL/CentOS-based servers
  sudo yum install python3 python3-pip -y
  

• Create a requirements.txt file in your project root listing all Python packages your script needs (include specific versions to avoid compatibility issues):

  pandas==2.1.4
  numpy==1.26.2
  # Add any other required packages here
  

• When deploying, navigate to your project directory on the server and run:

  pip3 install -r requirements.txt
  

• Pro tip: Avoid system-wide dependency conflicts by using a Python virtual environment:

  # Create a virtual environment
  python3 -m venv python-env
  # Activate it (Linux/Mac)
  source python-env/bin/activate
  # Windows: python-env\Scripts\activate
  
  # Install dependencies inside the env
  pip3 install -r requirements.txt
  

  Then point python-shell to the virtual environment’s Python executable in your Node.js code:

  const { PythonShell } = require('python-shell');
  
  PythonShell.run('your-script.py', {
    pythonPath: '/path/to/your/project/python-env/bin/python3'
  }, (err, results) => {
    if (err) throw err;
    // Process results here
  });
  

2. Containerized Deployment (Docker – Most Consistent Option) #

Docker eliminates "it works on my machine" issues by packaging your entire MEAN stack and Python environment into a single container:

• Create a Dockerfile in your project root:

  # Build the Node.js part first
  FROM node:18-alpine AS node-build
  WORKDIR /app
  COPY package*.json ./
  RUN npm install
  COPY . .
  
  # Set up the Python environment and combine with Node files
  FROM python:3.11-alpine
  WORKDIR /app
  # Copy Node.js app files from the previous build stage
  COPY --from=node-build /app .
  # Install Python dependencies
  COPY requirements.txt .
  RUN pip install --no-cache-dir -r requirements.txt
  
  # Start your Node.js app
  CMD ["npm", "start"]
  

• Don’t forget your requirements.txt file with all Python dependencies
• Build and run the container:

  docker build -t mean-python-app .
  docker run -p 3000:3000 mean-python-app
  

  This ensures your Python dependencies are installed exactly as needed, no matter where you deploy the container.

3. PaaS Platforms (e.g., Heroku, AWS Elastic Beanstalk) #

PaaS platforms have built-in tools to handle multi-language environments—here’s how to set them up:

Heroku #

• Make sure your project has both package.json (for Node.js dependencies) and requirements.txt (for Python)
• Add Heroku’s Node.js and Python buildpacks so the platform knows to set up both environments:

  heroku buildpacks:add heroku/nodejs
  heroku buildpacks:add heroku/python
  

  When you deploy, Heroku will automatically install Node.js dependencies, then read requirements.txt and install your Python packages.

AWS Elastic Beanstalk #

• Create a .ebextensions/python.config file in your project root to define Python setup steps:

  packages:
    yum:
      python3-devel: []
      python3-pip: []
  commands:
    01_install_python_deps:
      command: "pip3 install -r requirements.txt"
  

  When you deploy your app to Beanstalk, it will run these commands to install Python and your required packages before starting the Node.js app.

Key Things to Remember #

• Always specify exact version numbers in requirements.txt to avoid unexpected breaking changes from package updates
• If your Python script requires system-level dependencies (e.g., libpng for image processing), install them first—for example, in Docker use apk add libpng-dev, or on a VPS use sudo apt install libpng-dev
• Test your Python script directly in the deployment environment before starting the Node.js app to confirm all dependencies are working correctly

内容的提问来源于stack exchange，提问作者Abdelrahman AL-Sawalmeh