Skip to content

Run existing flows

The flows delivered by default are deployed by the administrators in each Prefect instance. They are listed in the Prefect flows section. This page shows how to run one of them from the Prefect dashboard, with the example of the Level-0 processing of a Sentinel-3 acquisition session.

1. Find the input data

The Level-0 processing takes as input an acquisition session of a CADIP station. To find the sessions available in the stations, use one of:

Note the session identifier, e.g. S3A_20260807020505053186.

2. Run the flow

  1. Open the Prefect dashboard, e.g. https://processing.ops.rs-python.eu/dashboard, and go to the Deployments page. Use the tags to filter the deployments, e.g. processing and l0.
  2. Open the process-l0/Process L0 deployment and click Run > Custom run.

  3. Fill in the parameters:

    • session: the session identifier found above.
    • env_args: set owner_id to your login, to run the flow on your behalf and write the products in your own collections.
    • force_systematic: leave it unchecked to only generate the Level-0 products. When checked, the next processings of the chain (e.g. the OLCI Level-1 and Level-2) are started automatically from the generated products, see Systematic processing.

    The other parameters (the processor, the Dask cluster, the target collections...) are the default settings of the platform, see Level-0 flows.

  4. Click Submit. A new flow run is created. See the Prefect documentation for more ways to run a deployment.

Tip

Use Quick run instead of Custom run to run a deployment with its default parameters.

3. Monitor the flow run

The flow run page shows:

  • the logs of the flow and of its tasks. The Level-0 flow first looks for the session in the catalog, stages it from the CADIP station if needed, then runs the Level-0 flow of the mission (process-s3-l0) as a sub-flow, which calls the generic processing flow (process-generic). The logs of the processing contain the link to the Dask dashboard of the processor run, and the command to cancel the run if needed.
  • the sub-flow runs and task runs with their status,
  • the artifacts, e.g. dpr-task-table and dpr-payload that show the processor task table and payload file.

Lower-level information (logs, traces, resources consumption) is available in Grafana, see Troubleshooting.

4. Look at the results

When the flow run is completed, the generated products are stored in the object storage and registered as STAC items in the catalog collections configured for the flow. Browse them with the catalog STAC browser, e.g. https://stac-browser-catalog.ops.rs-python.eu/, and download them as described in Browse products.

Going further