Building a Python + Node + pipenv + pm2 Docker Image and Automating Its Deployment

When developing Python applications, using pipenv for package management and pm2 from the Node ecosystem for process monitoring is recommended; this article walks through constructing a Docker image that bundles Python, Node, pipenv, and pm2.

The Dockerfile starts from the official python:3.7 base image, installs pipenv via pip3 , adds a stable Node v10.x runtime, installs pm2 with npm , and also installs vim for convenience.

To build the image, run: docker build -t python-node:1.0.0 . The trailing dot must point to the directory containing the Dockerfile.

After the image is built, push it to Docker Hub: set a tag such as wangying/python-node:latest (where wangying is the Docker Hub username and python-node the repository name), log in with docker login , and push the image. If the repository does not exist, Docker Hub creates it automatically.

To automate future builds, link Docker Hub to a GitHub account via Account Settings → Linked Accounts , create a GitHub repository (e.g., docker-images ), and configure the repository’s Build Settings on Docker Hub to use the GitHub source. Docker Hub will then trigger a build via GitHub’s webhook whenever the Dockerfile is updated.

When building a concrete application image on top of the base image, specify the base image, set a working directory, copy project files, expose the required port, install application dependencies, and define the container’s start command using pm2-runtime with a configuration file (e.g., pm2.json or the default ecosystem.config.js ).

The article concludes that the described steps solve the core problem of constructing and automating Docker images for Python‑Node environments.