Convert to using Dagster instead of Prefect and Celery

[abkfenris]

[abkfenris]

Programmatically creating jobs:

• https://stackoverflow.com/questions/72173805/submit-dagster-run-programmatically-without-dagit-graphql
• https://docs.dagster.io/_apidocs/libraries/dagster-graphql#dagster_graphql.DagsterGraphQLClient.submit_job_execution

Working with Django:

• https://stackoverflow.com/questions/59048161/integrating-dagster-with-django
• https://github.com/zkan/hello-dagster/blob/main/dagster-with-django-orm/hello_django_orm.py

[abkfenris]

Use a stator like system where instead enqueing runs directly from a subscription, toggle a 'ready to refresh' field at the dataset level.

Also have a default schedule field (hourly, 15 min, 5 min, subscription, server, defaults to hourly but as soon as it receives a subscription refresh gets switched), and a health check URL field.

Can start with just dataset level refreshing graph but can assemble upwards into a server refresh.

[abkfenris]

xref #837