Integrate new provenance.py script into operational deployment to tag products with Provenance information

[jordanpadams]

💡 Description

Follow-on to:

• pds-api 221: rework all/latest registry-api#193
• Add superseeded_by field in the opensearch index registry-mgr#63
• and ultimately, As a user, I want to query only the latest versions of products unless explicitly requested registry-api#441

Need to setup cron or lambda function or ??? to run on semi-regular cadence (every 45 minutes?) to update the operational indexes with the appropriate superseded_by metadata.

Rationale for 45 minute cadence:
Per discussion on NASA-PDS/registry-api#441, the script takes ~15 minutes to run on a few million products, so 45 minutes should give us plenty of leeway. Additionally, we can then tell node folks to expect their products to be life through the API ~60 minutes after they tag the products for release using Registry Manager.

[tloubrieu-jpl]

Steps to migrate:

1. export schema for each of the discipline nodes
2. recompute the new field values (here run provenance.py)
3. create a temporary index and add/remove node specificities
4. reindex the updated documents on the temporary index
5. drop the registry index
6. rename index

We need to re-index because the documents have been previously indexed (that sounds obvious, isn't it). What that does mean is: if the schema is updated that does not apply on document previously indexed.

This should be done in a script because that will need to be applied every build as we are expecting upgrades of the registry schema.

Overall we want to re-analyze how the provenance/all/latest feature is managed in the registry so to have an acceptable robustness/availability of the registries.

[jordanpadams]

@jimmie wrapping up documentation today. then heading onto schema updates

[jordanpadams]

all sub-tasks completed. demoed at 2/23 sprint review and a couple minor tweaks to be wrapped up today