NoSQL with MongoDB

MongoDB is the most popular NoSQL database. Instead of tables and rows, it stores data as flexible JSON-like documents in collections. This lesson covers MongoDB's document model, CRUD operations, and schema design patterns.

Documents and Collections

In MongoDB, data is organized into databases, collections, and documents:

Database:   myapp
  Collection: users
    Document: { _id: "...", name: "Alice", email: "a@mail.co" }
    Document: { _id: "...", name: "Bob", email: "b@mail.co" }
  Collection: posts
    Document: { _id: "...", title: "Hello World", author: "Alice" }

A document is a JSON object (technically BSON — Binary JSON). Unlike SQL rows, documents in the same collection can have different fields.

Setting Up

Install the MongoDB Node.js driver or Mongoose (an ODM — Object Document Mapper):

npm install mongoose

Connect to MongoDB:

Defining Schemas with Mongoose

Although MongoDB is schemaless, Mongoose lets you define schemas for validation and structure:

CRUD Operations

Create

Read

Query Operators

Operator Reference

Update

Delete

await User.findByIdAndDelete(userId);
await User.deleteOne({ email: "old@example.com" });
await User.deleteMany({ isActive: false });

Embedding vs Referencing

The biggest design decision in MongoDB is whether to embed related data or reference it.

Embedding (Denormalized)

Store related data inside the parent document:

Use embedding when:

• The related data is always accessed together
• The embedded array will not grow unboundedly
• The data does not need to be queried independently

Referencing (Normalized)

Store a reference (ID) to a document in another collection:

Use referencing when:

• The related data is large or accessed independently
• The related data changes frequently
• You need to query the related data on its own

Aggregation Pipeline

MongoDB's aggregation pipeline lets you process data in stages:

Practical Exercise

Build a blog schema with MongoDB:

Key Takeaways

• MongoDB stores data as flexible JSON-like documents grouped into collections.
• Mongoose adds schema validation and convenience methods on top of the native driver.
• Choose embedding for tightly coupled data and referencing for independent, large, or frequently changing data.
• Query operators like $gt, $in, $or, and $regex give you powerful filtering capabilities.
• The aggregation pipeline processes data in stages for complex analytics queries.