Skip to content

ACCESS-NRI/access-ram-condaenv

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

access-ram-condaenv

This repository is created taking payu-condaenv as a reference.

Overview

This repository is responsible for building, packaging and deploying the environment needed to run the replace_landsurface python scripts, as part of ACCESS-RAM3. This environment is created as a micromamba environment.

Usage

Triggering a Deployment

In order to trigger a deployment, the following steps must be followed:

  • Open a PR modifying the env.yml file.
  • The env.yml will be checked for validity.
  • When the PR is merged, the env.yml will be used to create a micromamba environment. This is then packaged using conda-pack, and deployed to the appropriate targets (eg. Gadi).

Using the Deployed Environment

On all of the deployment targets, the deployed environment can be activated using Environment Modules.

Gadi

  1. Make sure you're a member of the <PROJECT_TO_BE_DEFINED> project! If not, see how to join an NCI project.

    [!IMPORTANT] Make sure you do not have another conda environment active: either run conda deactivate or module unload any modules that are using conda.

  2. Once you are a member, run the following:

    module use /g/data/vk83/modules
    module load conda/access-ram3/<VERSION>

You are set to run ACCESS-RAM3.

Notes

On Future Deployment Environments

New deployment environments must be created as a GitHub Environment and also have an entry in the config/deployment-environment.json file.

Deploying locally

To deploy locally, you can use the assets created in the release. Releases are found here. Specifically:

  • To use the compressed environment (which doesn't require conda or python) you can run tar -xzf access-ram-<VERSION>.tar.gz access-ram and then ./access-ram/bin/activate to activate the environment.
  • To use the lockfile, you can run micromamba create -n environment-name -f access-ram.conda-lock.yml with an appropriate install of micromamba.