Reduce $lookup Operations

On this page

About this Task

Example
Learn More

The $lookup operator joins information from multiple collections into a single document. While the $lookup operation is useful when used infrequently, it can be slow and resource-intensive compared to operations that only query a single collection. If you often use $lookup operations, consider restructuring your schema to store related data in a single collection. This can improve query performance and reduce the cost of operations.

About this Task

Consider the following schema with two separate collections: products and orders. Each order can contain multiple products, and you want to keep track of product details within each order for quick access. The two separate collections are joined by a $lookup operation.

//products collection
db.products.insertMany( [
   {
      _id: 1,
      name: "Laptop",
      price: 1000,
      manufacturer: "TechCorp",
      category: "Electronics",
      description: "Fastest computer on the market."
   },
   {
      _id: 2,
      name: "Headphones",
      price: 100,
      manufacturer: "Sound",
      category: "Accessories",
      description: "The latest sound technology."
   },
   {
      _id: 3,
      name: "Tablet",
      price: 200,
      manufacturer: "TechCorp",
      category: "Electronics",
      description: "The most compact tablet."
   }
] )

//orders collection
db.orders.insertMany( [
   {
      _id: 101,
      customer_name: "John Doe",
      timestamp: "2024-05-11T010:00:00Z",
      product_ids: [1, 2],
      total: 1200
   },
   {
      _id: 102,
      customer_name: "Jane Smith",
      timestamp: "2024-05-11T012:00:00Z",
      product_ids: [2],
      total: 100
   }
] )

In this schema, you need to use the $lookup operation everytime you access order information. Running the $lookup operation adds query complexity and degrades performance. To reduce the use of $lookup operations, store data that is accessed together in a single collection .

Example

You can use the subset schema design pattern to embed a subset of product details in the orders collection. This lets you query a single collection to return the required results. Product details and documents not relevant to the orders collection remain in the products collection.

//orders collection
db.orders.insertMany( [
   {
      _id: 101,
      customer_name: "John Doe",
      timestamp: "2024-05-11T10:00:00Z",
      products: [
         {
            product_id: 1,
            name: "Laptop",
            price: 1000
         },
         {
            product_id: 2,
            name: "Headphones",
            price: 100
         }
      ],
      total: 1100
   },
   {
      _id: 102,
      customer_name: "Jane Smith",
      timestamp: "2024-05-11T12:00:00Z",
      products: [
         {
            product_id: 2,
            name: "Headphones",
            price: 100
         }
      ],
      total: 100
   }
] )

//products collection
db.products.insertMany( [
   {
      _id: 1,
      name: "Laptop",
      price: 1000,
      manufacturer: "TechCorp",
      category: "Electronics",
      description: "Fastest computer on the market."
   },
   {
      _id: 2,
      name: "Headphones",
      price: 100,
      manufacturer: "Sound",
      category: "Accessories",
      description: "The latest sound technology."
   },
   {
      _id: 3,
      name: "Tablet",
      price: 200,
      manufacturer: "TechCorp",
      category: "Electronics",
      description: "The most compact tablet."
   }
] )

This approach lets you keep collections separate while avoiding multiple queries by embedding key fields from the product collection into the orders collection. This improves read performance and simplifies data retrieval, as you can access all necessary information in a single query. However, it's important to consider potential document size limitations and data duplication.

Learn More

Back

Bloated Documents

Data Modeling Concepts