When you embed related data in a single document, you may duplicate data between two collections. Duplicating data lets your application query related information about multiple entities in a single query while logically separating entities in your model.
One concern with duplicating data is increased storage costs. However, the benefits of optimizing access patterns generally outweigh potential cost increases from storage.
Before you duplicate data, consider the following factors:
How often the duplicated data needs to be updated. Frequently updating duplicated data can cause heavy workloads and performance issues. However, the extra logic needed to handle infrequent updates is less costly than performing joins (lookups) on read operations.
The performance benefit for reads when data is duplicated. Duplicating data can remove the need to perform joins across multiple collections, which can improve application performance.
The following example shows how to duplicate data in an e-commerce application schema to improve data access and performance.
Create the following collections in the eCommerce database:
Collection Name
Description
Sample Document
customers
Stores customer information such as name, email, and phone number.
db.customers.insertOne( { customerId: 123, name: "Alexa Edwards", email: "a.edwards@randomEmail.com", phone: "202-555-0183"} )products
Stores product information such as price, size, and material.
db.products.insertOne( { productId: 456, product: "sweater", price: 30, size: "L", material: "silk", manufacturer: "Cool Clothes Co"} )orders
Stores order information such as date and total price. Documents in the orders collection embed the corresponding products for that order in the lineItems field.
db.orders.insertOne( { orderId: 789, customerId: 123, totalPrice: 45, date: ISODate("2023-05-22"), lineItems: [ { productId: 456, product: "sweater", price: 30, size: "L" }, { productId: 809, product: "t-shirt", price: 10, size: "M" }, { productId: 910, product: "socks", price: 5, size: "S" } ]} )The following properties from the products collection are duplicated in the orders collection:
productId
product
price
size
When the application displays order information, it displays the corresponding order's line items. If the order and product information were stored in separate collections, the application would need to perform a $lookup to join data from two collections. Lookup operations are often expensive and have poor performance.
The reason to duplicate product information as opposed to only embedding line items in the orders collection is that the application only needs a subset of product information when displaying orders. By only embedding the required fields, the application can store additional product details without adding unnecessary bloat to the orders collection.
The following example uses the subset pattern to optimize access patterns for an online store.
Consider an application where when user views a product, the application displays the product's information and five most recent reviews. The reviews are stored in both a products collection and a reviews collection.
When a new review is written, the following writes occur:
The review is inserted into the reviews collection.
The array of recent reviews in the products collection is updated with $pop and $push.
Create the following collections in the productsAndReviews database:
Collection Name
Description
Sample Document
products
Stores product information. Documents in the products collection embed the five most recent product reviews in the recentReviews field.
db.products.insertOne( { productId: 123, name: "laptop", price: 200, recentReviews: [ { reviewId: 456, author: "Pat Simon", stars: 4, comment: "Great for schoolwork", date: ISODate("2023-06-29") }, { reviewId: 789, author: "Edie Short", stars: 2, comment: "Not enough RAM", date: ISODate("2023-06-22") } ]} )reviews
Stores all reviews for products (not only recent reviews). Documents in the reviews collection contain a productId field that indicates the product that the review pertains to.
db.reviews.insertOne( { reviewId: 456, productId: 123, author: "Pat Simon", stars: 4, comment: "Great for schoolwork", date: ISODate("2023-06-29")} )The application only needs to make one call to the database to return the all information it needs to display. If data was stored entirely in separate collections, the application would need to join data from the products and reviews collection, which could cause performance issues.
Reviews are rarely updated, so it is not expensive to store duplicate data and keeping the data consistent between collections is not a challenge.
To learn how to keep duplicate data consistent, see Data Consistency.
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4