The best way to get accustomed with various data models offered by Cosmos DB would be to implement inherently relational domain models using the provided NoSQL data access APIs. This way, it is easier to grasp the benefits of different data models.
For this exercise, let's create a relational data model for our auction applications.
In this setup, we have three big clusters of data:
- Vehicles, which includes the manufacturer, model, year, engine specifications, and some additional attributes describing the car
- Users, which consists of the sellers and buyers of the cars sold through auctions
- Auctions, which consists of some metadata about the sale that's provided by the selling user, as well as the vehicles and bids provided by the users
We will describe this data using the SQL API.