ESM-VFC stacks

Central compute environment definition(s) for ESM VFC.

How to use?

Just use the environment file

You can just use the env file to install everything with Conda.

Note, however, that this env file does not specify any of the Jupyter and Dask configs that we also set in this repo.

Use the pre-built Docker images with Docker

You can use docker images that come with a fully working (repo2docker-built) Jupyter environment and run them with docker.

Assuming you want to expose your work directory to /work in the container, and run:

docker pull esmvfc/esm-vfc-stacks:latest
docker run \
    -v ${HOME}/work/:/work -w /work \
    -p 8888:8888 esmvfc/esm-vfc-stacks:latest \
    jupyter lab --no-browser --ip="0.0.0.0" --port="8888"

If you want to run a specific version of this image, replace the tag latest accordingly. All available versions are found on Dockerhub.

Use the pre-built Docker images with Singularity

To run from your work/ dir with Singularity:

cd ${HOME}/work/
singularity run \
    -B $(mktemp -d):/run/user \
    docker://esmvfc/esm-vfc-stacks:<tag> \
    jupyer lab --no-browser --ip="0.0.0.0"

If you want to run a specific version of this image, replace the tag latest accordingly. All available versions are found on Dockerhub.

How to contribute?

Modify the env file or any of the Binder config or Dask config with a pull request.

Releases

There are no Github-based releases. Instead, we use calver to build whenever something is pushed to master and tag it as YYYY.MM.DD-<Git SHA> on dockerhub.