Running images and mounting files

An image is the immutable snapshot (docker pull). A container is a running (or stopped) instance (docker run). A volume or bind mount is how you persist files when the container exits. Install Docker for your OS first: Docker setup.

1. Verify the engine

Pull a tiny image and run it once. --rm deletes the container when it exits, so it does not clutter docker ps -a:

docker run --rm hello-world

You should see a message that the client reached the Docker daemon and pulled the hello-world image. On Linux Engine installs that still require sudo, prefix the command with sudo until you are in the docker group.

2. List images and containers

docker images
docker ps -a

Run an image interactively and destroy it on exit:

docker run --rm -it ubuntu:24.04 bash

3. Mount a host directory

Mount a directory you own so work survives the container. Replace /path/on/host with that path:

docker run --rm -it -v /path/on/host:/work ubuntu:24.04 bash

Inside the container, files under /work are the same files as on the host. On Docker Desktop for Mac and Windows, prefer a path inside your home directory; bind mounts outside it can fail or be slow.

4. Stop and clean up

Stop a runaway container from another terminal with docker ps to get the ID, then docker stop with that ID. Remove unused images when disk fills:

docker image prune

Official command reference: docker run.

MESA and other research images

This site is the gateway: it teaches Docker, Python, Git, and workflow. It does not document how to run MESA.

When the astr400 MESA Docker repository is published, its usage docs will live in that repo and on Read the Docs, versioned with image tags. Link there for docker pull of the MESA image, work directories, and MESA commands. Link back here for installing Docker and for this generic docker run pattern.

Until that URL exists, use the MESA documentation for the code itself, and watch astr400 on GitHub for the image repository.