Step 2: Execution Environment

This is the second step of the RDMC Creation Workflow. You come here after you have added all your artifacts in Step 1 (continued): Upload and Describe an Artifact. Here you prepare a Reusable Execution Environment (REE) — a recording of the software environment needed to run your work, so that other people can reproduce your results later. When you are done, click Continue to go to Step 3: Container Metadata.

What Is a Reusable Execution Environment (REE)?

An REE is a separate, replicable context that captures everything needed to run your software: the source code, the runtime it was built with, and a list of the software it depends on. The REE is prepared on a dedicated REE service and the results are saved back into your container, so the environment can be rebuilt and validated independently of your research artifacts.

Note

If the Execution Environment feature is disabled in your configuration, this page only shows a short notice. You can simply click Continue and come back later.

The REE Workbench

This step opens the REE Workbench. The first time you visit it, you are asked to give your Execution Environment a name:

  1. Enter a name for your Execution Environment.

  2. Click Create.

The workbench then opens with three areas: Components, Pipeline, and Saved to Container. Your environment stays linked to this container — when you leave and come back, the workbench reopens the same environment. You can rename it at any time with the pencil icon next to its name.

Components

An Execution Environment is assembled from three component slots:

  • Repository — an archive of your source code (a .zip of your project). It must contain an apptainer.def recipe at its top level — this is what the Build job turns into a container image.

  • Runtime — a ready-made container image. It can be an image made elsewhere, or a build result you saved from an earlier run (see Saved to Container below). The Validate job runs this image.

  • SBOM — the Software Bill of Materials, a machine-readable list (.json) of the software your work depends on.

Slots are filled with Pick from container: choose a file you already uploaded as a resource in Step 1. Use the artifact and folder drop-downs above the slots to narrow down where to look. For the SBOM slot, you can alternatively use Upload from device to upload a new file directly from your computer; the file is then also kept as a resource of your container.

Each slot only accepts certain file types; if a file is not accepted, the workbench shows the exact reason reported by the REE service. You can clear a slot and fill it again at any time.

Pipeline

Once the required components are in place, the Pipeline panel lets you run jobs on the REE service:

  • Export — packs the current components into a single archive. Builds and validations always run from an export, so run it first — and again after adding or changing components.

  • Build — turns the repository’s apptainer.def into a container image, using the latest export as input.

  • Validate — runs the runtime image to check that the environment actually works; whatever the image writes to /output comes back as an artifact.

The panel shows which jobs are ready to start and the status of jobs that are running or finished. If you change a component after a job has run, its result is marked outdated — the artifact stays downloadable, but you should re-export before running a job against the new selection. For every job, you can:

  • open or download its log to see what happened,

  • download the resulting artifact (for example, the built environment),

  • save the artifact into your container (see below).

The pipeline keeps the last five runs per job type; older runs, their logs, and their artifacts are deleted automatically. Save a result to the container if you want to keep it.

Saved to Container

Job results become part of your research data only when you save them into the container. When you click Save to container on a finished job, you choose the artifact (and optionally a folder) the result should be stored in. The file then appears:

  • in the Saved to Container panel of the workbench, and

  • as a regular resource of your container in Step 1: Artifact Initialization, so it is included when the container is sealed.

Each saved entry shows its descriptive title (for example, REE Build Result (Job #28)) together with the stored file name, so you can match it against the component slots. If a saved file is currently selected as a component, its row shows a set as … marker.

If you saved something by mistake, you can remove it again from the Saved to Container panel.

Note

Removing a saved file from the container does not clear the component slot that uses it — the execution environment keeps its own copy of every selected component. If you no longer want the file as a component, clear or change the slot as well.

When you are done, click Continue to go to Step 3: Container Metadata.