NOMAD Oasis Distribution Template

This repository is a template for creating your own custom NOMAD Oasis distribution image. Click here to use this template, or click the Use this template button in the upper right corner of the main GitHub page for this template.

Important

The templated repository will run a GitHub action on creation which might take a few minutes. After the workflow finishes you should refresh the page and this message should disappear. If this message persists you might need to trigger the workflow manually by navigating to the "Actions" tab at the top, clicking "Template Repository Initialization" on the left side, and triggering it by clicking "Run workflow" under the "Run workflow" button on the right.

FAIRmat-NFDI's NOMAD Oasis Distribution

This is the NOMAD Oasis distribution of FAIRmat-NFDI. Below are instructions for how to deploy this distribution and how to customize it through adding plugins.

Important

Depending on the settings of the owner of this repository, the distributed image might be private and require authentication to pull. If you want to keep the image private you need to configure and use a personal access token (PAT) according to the instructions in the GitHub docs here. If you want to make the image public (recommended), you should make sure that your organization settings allow public packages and make this package public after building it. You can read more about this in the GitHub docs here.

Tip

In order for others to find and learn from your distribution we in FAIRmat would greatly appreciate it if you would add the topic nomad-distribution by clicking the ⚙️ next to "About" on the main GitHub page for this repository.

In this README you will find instructions for:

1. Deploying the distribution
2. Adding a plugin
3. Using the jupyter image
4. Updating the distribution from the template
5. Solving common issues

Deploying the distribution

Below are instructions for how to deploy this NOMAD Oasis distribution for a new Oasis and for an existing Oasis

For a new Oasis

1. Make sure you have docker installed. Docker nowadays comes with docker compose built in. Prior, you needed to install the stand-alone docker-compose.

2. Clone the repository or download the repository as a zip file.

   git clone https://github.com/FAIRmat-NFDI/nomad-distro-template.git
   cd nomad-distro-template

   or

   curl-L -o nomad-distro-template.zip "https://github.com/FAIRmat-NFDI/nomad-distro-template/archive/main.zip"
   unzip nomad-distro-template.zip
   cd nomad-distro-template

3. On Linux only, recursively change the owner of the .volumes directory to the nomad user (1000)

   sudo chown -R 1000 .volumes

4. Pull the images specified in the docker-compose.yaml

   Note that the image needs to be public or you need to provide a PAT (see "Important" note above).

   docker compose pull

5. And run it with docker compose in detached (--detach or -d) mode

   docker compose up -d

6. Optionally you can now test that NOMAD is running with

   curl localhost/nomad-oasis/alive
   

7. Finally, open http://localhost/nomad-oasis in your browser to start using your new NOMAD Oasis.

Updating the image

1. Whenever you update your image you need to shut down NOMAD using

   docker compose down

   and then repeat steps 4. and 5. above.

2. You can remove unused images to free up space by running

   docker image prune -a

NOMAD Remote Tools Hub (NORTH)

To run NORTH (the NOMAD Remote Tools Hub), the hub container needs to run docker and the container has to be run under the docker group. You need to replace the default group id 991 in the docker-compose.yaml's hub section with your systems docker group id. Run id if you are a docker user, or getent group | grep docker to find your systems docker gid. The user id 1000 is used as the nomad user inside all containers.

Please see the Jupyter image section below for more information on the jupyter NORTH image being generated in this repository.

You can find more details on setting up and maintaining an Oasis in the NOMAD docs here: nomad-lab.eu/prod/v1/docs/oasis/install.html

For an existing Oasis

If you already have an Oasis running you only need to change the image being pulled in your docker-compose.yaml with ghcr.io/fairmat-nfdi/nomad-distro-template:main for the services worker, app, north, and logtransfer.

If you want to use the nomad.yaml from this repository you also need to comment out the inclusion of the nomad.yaml under the volumes key of those services in the docker-compose.yaml.

volumes:
  # - ./configs/nomad.yaml:/app/nomad.yaml

To run the new image you can follow steps 5. and 6. above.

Adding a plugin

To add a new plugin to the docker image you should add it to the plugins table in the pyproject.toml file.

Here you can put either plugins distributed to PyPI, e.g.

[project.optional-dependencies]
plugins = [
  "nomad-material-processing>=1.0.0",
]

or plugins in a git repository with either the commit hash

[project.optional-dependencies]
plugins = [
  "nomad-measurements @ git+https://github.com/FAIRmat-NFDI/nomad-measurements.git@71b7e8c9bb376ce9e8610aba9a20be0b5bce6775",
]

or with a tag

[project.optional-dependencies]
plugins = [
  "nomad-measurements @ git+https://github.com/FAIRmat-NFDI/[email protected]"
]

To add a plugin in a subdirectory of a git repository you can use the subdirectory option, e.g.

[project.optional-dependencies]
plugins = [
  "ikz_pld_plugin @ git+https://github.com/FAIRmat-NFDI/AreaA-data_modeling_and_schemas.git@30fc90843428d1b36a1d222874803abae8b1cb42#subdirectory=PVD/PLD/jeremy_ikz/ikz_pld_plugin"
]

Once the changes have been committed to the main branch, the new image will automatically be generated.

The Jupyter image

In addition to the Docker image for running the oasis, this repository also builds a custom NORTH image for running a jupyter hub with the installed plugins. This image has been added to the configs/nomad.yaml during the initialization of this repository and should therefore already be available in your Oasis under "Analyze / NOMAD Remote Tools Hub / jupyter"

The image is quite large and might cause a timeout the first time it is run. In order to avoid this you can pre pull the image with:

docker pull ghcr.io/fairmat-nfdi/nomad-distro-template/jupyter:main

If you want additional python packages to be available to all users in the jupyter hub you can add those to the jupyter table in the pyproject.toml:

[project.optional-dependencies]
jupyter = [
  "voila",
  "ipyaggrid",
  "ipysheet",
  "ipydatagrid",
  "jupyter-flex",
]

Updating the distribution from the template

In order to update an existing distribution with any potential changes in the template you can add a new git remote for the template and merge with that one while allowing for unrelated histories:

git remote add template https://github.com/FAIRmat-NFDI/nomad-distro-template
git fetch template
git merge template/main --allow-unrelated-histories

Most likely this will result in some merge conflicts which will need to be resolved. At the very least the Dockerfile and GitHub workflows should be taken from "theirs":

git checkout --theirs Dockerfile
git checkout --theirs .github/workflows/docker-publish.yml

For detailed instructions on how to resolve the merge conflicts between different version we refer you to the latest template release notes

Once the merge conflicts are resolved you should add the changes and commit them

git add -A
git commit -m "Updated to new distribution version"

Ideally all workflows should be triggered automatically but you might need to run the initialization one manually by navigating to the "Actions" tab at the top, clicking "Template Repository Initialization" on the left side, and triggering it by clicking "Run workflow" under the "Run workflow" button on the right.

FAQ/Trouble shooting

I get an Error response from daemon: Head "https://ghcr.io/v2/FAIRmat-NFDI/nomad-distro-template/manifests/main": unauthorized when trying to pull my docker image.

Most likely you have not made the package public or provided a personal access token (PAT). You can read how to make your package public in the GitHub docs here or how to configure a PAT (if you want to keep the distribution private) in the GitHub docs here.