Remote Development on Docker Containers

Provision Docker containers as Coder workspaces with this example template.

Prerequisites

Infrastructure

Running Coder inside Docker

If you installed Coder as a container within Docker, you will have to do the following things:

Make the the Docker socket available to the container
- (recommended) Mount /var/run/docker.sock via --mount/volume
- (advanced) Restrict the Docker socket via https://github.com/Tecnativa/docker-socket-proxy
Set --group-add/group_add to the GID of the Docker group on the host machine
- You can get the GID by running getent group docker on the host machine

If you are using docker-compose, here is an example on how to do those things (don't forget to edit group_add!): 0bfe0d63ae/docker-compose.yaml (L16-L23)

Running Coder outside of Docker

If you installed Coder as a system package, the VM you run Coder on must have a running Docker socket and the coder user must be added to the Docker group:

# Add coder user to Docker group
sudo adduser coder docker

# Restart Coder server
sudo systemctl restart coder

# Test Docker
sudo -u coder docker ps

Architecture

This template provisions the following resources:

Docker image (built by Docker socket and kept locally)
Docker container pod (ephemeral)
Docker volume (persistent on /home/coder)

This means, when the workspace restarts, any tools or files outside of the home directory are not persisted. To pre-bake tools into the workspace (e.g. python3), modify the container image. Alternatively, individual developers can personalize their workspaces with dotfiles.

Note

This template is designed to be a starting point! Edit the Terraform to extend the template to support your use case.

Editing the image

Edit the Dockerfile and run coder templates push to update workspaces.

2.2 KiB Raw Blame History