Data view caching

[SamStudio8]

With the desire to EOL the elan metadata file, it would be best to have the API cache any queries that return very commonly requested data before spawning a celery task to do the same thing all over again. The good news is we can do it with the PAG entry point so that aligns with the plan to cache at the PAG level for now.

[SamStudio8]

In COG parlance, we basically need an MDV that can grab all the C-class data, and it shouldn't take 20 minutes.