How to deploy a Django application to any VPS in 20 minutes with Docker, Kamal and SaaS Pegasus

Summary

Cory Zue demonstrates deploying a Django application to a VPS using SaaS Pegasus, Docker, and Kamal. The tutorial uses a DigitalOcean droplet, while noting that VPS hosting is available from many providers, including Linode and AWS EC2.

Deployment walkthrough

The tutorial covers how to:

• Create an Ubuntu 22.04 droplet with 2 GB of memory.
• Install Kamal on the developer’s local machine and Docker on the server.
• Set up a deployment user, grant it Docker access, and configure SSH-key login.
• Create a private Docker Hub repository and access token for storing the application image.
• Update Kamal’s deployment configuration with the image repository, server IP, registry credentials, and application secrets, including the PostgreSQL password and Django secret key.
• Use the server’s IP address rather than configuring a domain. The presenter notes that DNS endpoints can be used for a domain and SSL setup.

What Kamal sets up

Running kamal setup builds the application’s Docker image, including frontend assets and Python packages, pushes it to Docker Hub, and then pulls it onto the VPS. It starts the Django app and supporting services, including PostgreSQL, Redis, Traefik, and Celery, and checks the application’s health.

The deployment completes in about six minutes in the demonstration. The presenter then opens the app in a browser and confirms that account creation works, describing the full process as taking under 20 minutes.

Content type: Hands-on deployment tutorial; no product review is presented.

Main speaker/source: Cory Zue, from SaaS Pegasus.