Data Analytics using MongoDB

I am learning Mongodb for a project in my company which is a chat process. Chat process will show some analytics based on sales and purchase history data.

I have found mongo db difficult as compared to MSSQL to build queries and to generate dataset / result from multiple collections with multiple filters.
So, I have few questions if someone can help me to clear.

Can we use Mongodb for large scale data storage?
Can we use Mongodb data for analytics?
Is it possible to create stored procedures like in MSSQL in mongodb where we can pass different parameters and generate some output?

Please help me to find answers for the above.

Naveen Jain

Hey @Naveen_72857

Yep, extremely large datasets as mongodb offers sharding which is spliting a collection over multiple servers.

Yes there is no reason why not.

I do not know about SQL besides that it exists. However this sounds like something you would do at an application level with a function and passing different params to that function which will inturn query mongodb

Hi @Naveen_72857,

Have you tried aggregations ?

As you know, MongoDB is a distributed database which has been designed for scalability. It can easily store large scale datasets.

Yes, we can definitely do that. The main question is what kind of data you are having and the tools that you are using for analytics.

You can Store a JavaScript Function on the Server for reuse.

That being said, I would highly recommend you to consult an expert on this to get any professional advice.

Shubham Ranjan
Curriculum Support Engineer

My 2 cents:

If you’re thinking about the traditional data warehouse using Inmon or Kimball where you’re looking at different timelines of historic data, the structures are different and I don’t think MongoDB directly supports these methodologies. For example, the concept of a Slow Changing Dimension can be somewhat implemented using the Schema Versioning pattern or a combination of the Shema Versioning pattern and the Polymorphic Pattern. Traditional DWs is very much about relationships and MongoDB isn’t necessarily about that. That said, being a “schemaless” db, I think MongoDB will thrive in an Analytics world where requirements are fast changing (i.e. there’s a need to add/remove fields) and for fast reads (MongoDB aggregation pipeline, sharding, data models and config). In addition, I think MongoDB will do well when it comes to Real Time analytics if you use Atlas and they’ve also recently launched Charts in the Atlas Cluster for visualisation. Also look up the MongoDB BI Connector.

I’m sure that there are some Analytics Use Cases that the Curriculum Engineers can share (@Shubham_Ranjan)?

1 Like