Migrations: Implement bookkeeping so database tells us which schema it conforms to

[eecavanna]

In other words, the database would "know" (tell us) what migrations have been applied to it; or at least which one was applied to it most recently.

[eecavanna]

I plan to implement this as part of the standard code run by the Python notebooks used for migration. The notebooks could create a document at the beginning of the migration process; something like: { "from": __, "to": __, "start_time": "...", "end_time": null }. Then, at the end of the migration process, the notebook could update the document to have a non-null end_time. To determine the current schema version, a user could sort the documents by their end_time (the to value of the document having the latest end_time would be the schema version the database is "at").

Alternatively (or in addition), the notebook could also update some single/global nmdc_schema_version field to indicate the schema version to which the database was updated most recently. Then, a user could check that single nmdc_schema_version field as a quick way of checking (effectively) what schema version the database is "at".

[eecavanna]

Sprint planning note: There's a chance that this will be done on Friday afternoon, but I think it's more likely that it will spill over into next sprint.

[ssarrafan]

Moving to next sprint @eecavanna