For a topic, we may need to configure how handle it in data ingestion and UI tools.
Here are some examples:
1. pre-configure Kibana dashboards
2. date mapping in ES
3. Druid Kafka indexing service supervisors

These info will be saved in a db table, with such schema:
id
topic_id (FK to topic table)
content (TEXT type)
type_id (Kibana dashboard, ES mapping, Druid supervisor, etc) (FK to type table)

typ table:
id
db_id ((FK to db table)
name

At deployment time, pre-packed settings will be applied.

Admin UI should be able to view, edit and deploy such settings.