Data Modelling approach for MongoDB NoSQL

I am an experienced business analyst just using MongoDB for the first time. It has raised a question in my mind.

Typically I user Entity Modelling or UML Class Diagrams to model data conceptually/logically before passing on my designs to the DBA to be implemented physically.

Since with a MongoDB NoSQL database the physical implementation is different, it occurs to me there might be alternative paradigms to do the modelling/visualisation? Does anyone have any ideas or references to point me at?

Thanks

Ged

Hello @Gerald_Dunn, Welcome to the MongoDB Community forum!

In general, the entity models showing relationships is a good starting point in the data modeling process. While the relationships between entities remain same (one-to-one, one-to-many, many-to-many), the way you model for a physical database structure for MongoDB NoSQL database differs, most of the times.

MongoDB data allows storing data in a de-normalzed form. That is, for example, a typical Order and Order Details are stored together in a single collection (collection is analogous to a table in relational database). The main reason is that the data accessed together, is stored together. MongoDB has tools to work with such data, efficiently.

That said, the best place to start with is MongoDB data modeling is (and you will find links to related topics from there):

2 Likes

Thanks @Prasad_Saya …makes sense. I will look at that documentation and then share my thoughts!

Any other thoughts from anyone would be gratefully received!