MongoDB Comprehensive Data Modeling Guide

📦 Model for Workloads

Relational to Document Model

Transitioning from relational to document-oriented databases involves understanding how data can be embedded or referenced.

Example Transition

// Relational
User Table: user_id, name
Order Table: order_id, user_id, product

// Document Model
{
  user_id: 1,
  name: "Alex",
  orders: [{ order_id: 1001, product: "Book" }]
}
    

Identifying and Quantifying Entities

Analyze your domain to clearly identify unique entities like users, orders, products, and quantify them in terms of document size and frequency.

Identifying Reads and Writes

Determine which data is frequently read or written to decide how to structure your documents efficiently.

Quantifying Read and Write

Measure your application's performance and frequency of operations to make informed modeling decisions.

🔗 Design Relationships

Identifying Relationships

Understand clearly if relationships are 1:1, 1:N, or N:N.

Embedding or Referencing

Embed data for fast reads; reference data if independence or size is a concern.

Modeling One-to-One

// Embedded (preferred)
{
  user: "Alex",
  profile: { age: 30, gender: "male" }
}
    

Modeling One-to-Many

// Reference
user: { user_id: 1, name: "Alex" }
orders: { order_id: 10, user_id: 1, product: "Book" }
    

Modeling Many-to-Many

// Relationship collection
user_courses: { user_id: 1, course_id: 5 }
    

✅ Validate Schemas

Schema Validation

MongoDB allows schema validation using JSON Schema to ensure data integrity.

Code Summary: Schema Validation

db.createCollection("sales", {
  validator: {
    "$and": [
      { "$expr": { "$lt": ["$items.discountedPrice", "$items.price"] } },
      { "$jsonSchema": {"properties": {"items": {"bsonType": "array"}}}}
    ]
  }
});
    

This ensures discountedPrice is always lower than price and validates items as an array.

Finding Non-matching Documents

db.collection.find({ $nor: [validator] });

Real-World Example: Bookstore Reviews Validator

const bookstore_reviews_default = { /* schema definition here */ };
const schema_validation = { $jsonSchema: bookstore_reviews_default };

db.runCommand({
  collMod: "reviews",
  validator: schema_validation,
  validationLevel: "strict",
  validationAction: "error"
});
    

This code enforces strict validation, ensuring documents match the schema exactly.