Determine how job output will be returned

[manning-ncsa]

If the jobs will output files, they can be be served upon job completion following the pattern in https://gitlab.com/spt3g/kubernetes/-/tree/main/src/spt-api-server/, where a dedicated, lightweight fileserver is deployed to serve job output files, where the URL for download is based on the job UUID. This could be utilized by the frontend web app via an AJAX request that fetches and renders the output files in the browser, and/or directly as a simple download link the user opens manually when the job completes.

If the jobs will output a small enough amount of textual data, then perhaps job output could be stored as a JSON blob in a dedicated database table associated with the job table (where the foreign key is the job UUID).