Deploying on Vercel and Netlify

Vercel and Netfliy are popular hosted platforms for deploying static website.

They make it easy and convenient to host static websites from existing git repositories, and make them widely available via their CDN.

Netlify

To deploy your own JupyterLite on Netlify, you can start from the JupyterLite Demo by generating a new repository from the template.

Then add a runtime.txt file with 3.10 as the content to specify Python 3.10 as dependency.

Finally specify jupyter lite build --contents content --output-dir dist as the “Build Command”, and dist as “Published Directory”:

You might also want to specify the --debug flag to get extra log messages:

Vercel

Just like Netlify, Vercel can connect to an existing git repository and seamlessly deploy static files on push and PR events (previews).

Here the configuration is very similar to Netlify. You can specify the same jupyter lite build --contents content --output-dir dist build command and configure dist as the published directory.

Using micromamba to create the build environment

The build environments of hosted platforms like Netlify and Vercel generally allow for limited control on the Python version installed on the build machine.

This was the case for a while when their build image only included Python 3.6 while JupyterLite requires Python 3.10+. This can be limiting in some cases, especially when you want to have more control on the build.

Fortunately it is possible to run arbitrary bash scripts, which provides a convenient escape hatch.

Specify the Python packages in a requirements-deploy.txt file with additional dependencies if needed:

jupyterlab~=3.4
jupyterlite-core
jupyterlite-pyodide-kernel

Then create a new deploy.sh file with the following content:

#!/bin/bash
set -e

# Install wget
# yum is not available on netlify, but it´s not a problem because wget is already installed, this validation is just to avoid errors on build step.
if command -v yum &> /dev/null; then
    yum install wget -y
fi

# Download and extract Micromamba
wget -qO- https://micro.mamba.pm/api/micromamba/linux-64/latest | tar -xvj bin/micromamba

# Set up environment variables
export MAMBA_ROOT_PREFIX="$PWD/micromamba"
export PATH="$PWD/bin:$PATH"

# Create the environment
micromamba create -n jupyterenv python=3.12 -c conda-forge -y

# Install dependencies via pip in the micromamba environment
micromamba run -n jupyterenv python -m pip install -r requirements-deploy.txt

# Build JupyterLite
micromamba run -n jupyterenv jupyter lite --version
micromamba run -n jupyterenv jupyter lite build --contents content --output-dir dist

Micromamba creates a new self-contained environment, which makes it very convenient to install any required package without being limited by the build image.

Then configure the build command and output directory, for example on Vercel:

You might also want to specify the --debug flag to get extra log messages:

jupyter lite build --debug

Configuring HTTP headers

Starting with JupyterLite 0.4.0, it is possible to have more reliable file system access from the kernel via the use of synchronous communication with the kernel over the SharedArrayBuffer browser feature. See the Contents guide for more information.

However this requires setting two HTTP headers for the SharedArrayBuffer to be enabled on the page.

Adding the headers to Netlify

Create a netlify.toml file at the root of the repo with the following content:

[[headers]]
for = "/*"
  [headers.values]
    Cross-Origin-Opener-Policy = "same-origin"
    Cross-Origin-Embedder-Policy = "require-corp"

Adding the headers to Vercel

Create a vercel.json file at the root of the repo with the following content:

{
  "headers": [
    {
      "source": "/(.*)",
      "headers": [
        {
          "key": "Cross-Origin-Embedder-Policy",
          "value": "require-corp"
        },
        {
          "key": "Cross-Origin-Opener-Policy",
          "value": "same-origin"
        }
      ]
    }
  ]
}