EWC Community Hub Items Catalog

This repository hosts the official catalog of items offered in the EWC Community Hub.

Item's Metadata

Below we show an excerpt from items.yaml, to exemplify the metadata of an onboarded item. For more details on required attributes, as well as optional ones (such as inputs specification for ewccli compatibility), please refer to the items' metadata schema definition.

ecmwf-data-flavour:
  annotations:
    technology: "Ansible Playbook"
    category: "Compute,Data Access"
    supportLevel: "EWC"
    licenseType: "Apache License 2.0"
    others: "Deployable,EWCCLI-compatible"
  description: |
    Includes the basic ECMWF software stack, with MARS client and an environment with [ecCodes](https://github.com/ecmwf/eccodes), [Metview](https://metview.readthedocs.io/en/latest/index.html), [Earthkit](https://github.com/ecmwf/earthkit) and [Aviso](https://pyaviso.readthedocs.io/en/latest/).

    ## Usage
    Example usage:

      ```bash
      ansible-playbook -i inventory ecmwf-data-flavour.yml
      ```

    ## Inputs
    You may use the following ansible variables to customize this playbook:

    - `reboot_if_required`: Boolean to reboot the instance if required after an update. Default: `true`
    - `ecmwf_toolbox_env_wipe`: Boolean to decide whether to wipe the environment if exists prior to a reinstallation. Default: `false`
    - `ecmwf_toolbox_env_name`: Name of the environment containing the ECMWF toolbox. Default: `ecmwf-toolbox`
    - `ecmwf_toolbox_create_ipykernel`: Boolean to create a system-wide kernel available. Default: `true`
    - `conda_prefix`: Prefix where conda is installed. Default: `/opt/conda`
    - `conda_user`: User owning the conda installation. Default: `root`

  displayName: ECMWF Data Flavour
  ewccli:
    inputs:
      - name: reboot_if_required
        default: true
        description: Boolean to reboot the instance if required after an update.
        type: bool
      - name: ecmwf_toolbox_env_wipe
        default: false
        description: Boolean to decide whether to wipe the environment if exists prior to a reinstallation.
        type: bool
      - name: ecmwf_toolbox_env_name
        default: "ecmwf-toolbox"
        description: Name of the environment containing the ECMWF toolbox.
        type: str
      - name: ecmwf_toolbox_create_ipykernel
        default: true
        description: Boolean to create a system-wide kernel available.
        type: bool
      - name: conda_prefix
        default: "/opt/conda"
        description: Prefix where conda is installed.
        type: str
      - name: conda_user
        default: "root"
        description: User owning the conda installation.
        type: str
    pathToMainFile: playbooks/ecmwf-data-flavour/ecmwf-data-flavour.yml
    pathToRequirementsFile: requirements.yml
  home: https://github.com/ewcloud/ewc-flavours/tree/2.0.0/playbooks/ecmwf-data-flavour
  icon: https://raw.githubusercontent.com/ewcloud/ewc-community-hub/refs/heads/main/logos/EWCLogo.png
  license: https://github.com/ewcloud/ewc-flavours/blob/2.0.0/LICENSE
  maintainers:
    - name: EWC Team
      email: support@europeanweather.cloud
      url: https://github.com/ewcloud/ewc-flavours/issues
  name: "ecmwf-data-flavour"
  published: true
  sources:
    - https://github.com/ewcloud/ewc-flavours.git
  summary: Includes the basic ECMWF software stack, with MARS client and an environment with ecCodes, Metview, Earthkit and Aviso.
  version: "2.0.0"

⛔ The attribute name within each item entry must always match the key under which the item's metadata is defined, to enforce uniqueness of all entries.

⚠️ At least one of the maintainers[*].email or the maintainers[*].url attributes should be set. These is to ensure end-users can submit inquiries or receive support, in accordance with the support level offered by Item owners.

Schema Validation

💡 To learn more about how you can onboard your item into the catalog, please check the official EWC documentation.

This repository relies on GitHub actions to automate the process of catalog/item metadata validation. The pipelines are configured to run upon pull request opening, subject to approval of the maintainers.

Running Locally

If you wish to validate changes to the metadata on before opening a pull request, you can emulate the steps performed by the GitHub automation. Make sure your working environment has Docker installed to setup the validation tool locally (one time operation):

docker build --tag ewc/ajv-cli:5.0.0 .

Then, to validate any changes in the metadata against the expected schema, run:

docker run --rm --volume .:/tmp:ro \
  ewc/ajv-cli:5.0.0 \
  -s tmp/schemas/items/v1alpha1.json \
  -d /tmp/items.yaml \
  -c ajv-formats \
  --spec draft2020

If changes comply, you should see a successful run message like:

/tmp/items.yaml valid