Relations in Prisma in NestJS - Performance & Optimization

Performance: Relations in Prisma

MEDIUM IMPACT

This affects database query speed and server response time when fetching related data.

Fetching related data with Prisma relations

NestJS

const users = await prisma.user.findMany({ include: { posts: true } });

Fetches users and their posts in a single query, reducing database calls.

📈 Performance GainSingle database query regardless of number of users, greatly improving response time.

Fetching related data with Prisma relations

NestJS

const users = await prisma.user.findMany();
for (const user of users) {
  user.posts = await prisma.post.findMany({ where: { authorId: user.id } });
}

This triggers one query per user, causing many database calls and slow response.

📉 Performance CostTriggers N database queries for N users, increasing server response time linearly.

Performance Comparison

Pattern	Database Queries	Server Response Time	Client Impact	Verdict
N+1 Query Pattern	N+1 queries	High latency	Slow interaction response	[X] Bad
Single Query with include	1 query	Low latency	Fast interaction response	[OK] Good

Rendering Pipeline

Relations in Prisma affect the server-side data fetching stage before rendering. Efficient queries reduce server processing time and speed up data delivery to the frontend.

→Data Fetching

→Server Response

→Client Rendering

⚠️ BottleneckMultiple database queries increase server response time and delay rendering.

Core Web Vital Affected

INP

This affects database query speed and server response time when fetching related data.

Optimization Tips

1Avoid fetching related data inside loops to prevent many database queries.

2Use Prisma's include or select to fetch relations in a single query.

3Reducing database queries improves server response and user interaction speed.

Performance Quiz - 3 Questions

What is the main performance issue with fetching related data in a loop with Prisma?

A It causes many database queries, increasing server response time.

B It reduces database queries to one, improving speed.

C It caches data on the client side automatically.

D It improves browser rendering speed.

DevTools: Network

How to check: Open DevTools, go to Network tab, filter XHR requests, and observe the number of API calls made when loading related data.

What to look for: Multiple repeated API calls indicate N+1 query problem; a single call with all data is optimal.

Practice

1. What type of relation in Prisma connects one record to many records in another model?

easy

A. One-to-many relation

B. One-to-one relation

C. Many-to-many relation

D. No relation

Solution

1. Step 1: Understand relation types in Prisma

   Prisma supports one-to-one, one-to-many, and many-to-many relations to connect models.

2. Step 2: Identify the relation connecting one record to many

   One-to-many means one record in a model relates to multiple records in another model.

3. Final Answer:

   One-to-many relation -> Option A

4. Quick Check:

   One record linked to many = One-to-many [OK]

Hint: One record to many records means one-to-many relation [OK]

Common Mistakes:

• Confusing one-to-one with one-to-many
• Thinking many-to-many is one-to-many
• Assuming no relation means one-to-many

2. Which of the following is the correct Prisma schema syntax for a one-to-one relation between User and Profile models?

easy

A. model User {\n id Int @id @default(autoincrement())\n profiles Profile[]\n}\n\nmodel Profile {\n id Int @id @default(autoincrement())\n user User\n}

B. model User {\n id Int @id @default(autoincrement())\n profile Profile? @relation(fields: [profileId], references: [id])\n profileId Int?\n}\n\nmodel Profile {\n id Int @id @default(autoincrement())\n user User?\n}

C. model User {\n id Int @id @default(autoincrement())\n profile Profile\n}\n\nmodel Profile {\n id Int @id @default(autoincrement())\n user User\n}

D. model User {\n id Int @id @default(autoincrement())\n profileId Int\n}\n\nmodel Profile {\n id Int @id @default(autoincrement())\n}

Solution

1. Step 1: Check one-to-one relation syntax

   One-to-one relations use a single optional foreign key field with @relation specifying fields and references.

2. Step 2: Verify the syntax

   The correct schema declares profile Profile? @relation(fields: [profileId], references: [id]) and profileId Int? in User, with user User? in Profile. This matches the pattern for optional one-to-one.

3. Final Answer:

   Schema with explicit optional FK and @relation -> Option B

4. Quick Check:

   One-to-one uses single foreign key with @relation [OK]

Hint: One-to-one needs single foreign key with @relation fields [OK]

Common Mistakes:

• Using array [] for one-to-one relation
• Missing @relation fields and references
• Not making foreign key optional

3. Given the Prisma models below, what will be the result of the query prisma.user.findUnique({ where: { id: 1 }, include: { posts: true } }) if user with id 1 has two posts?

model User {
  id    Int    @id @default(autoincrement())
  name  String
  posts Post[]
}

model Post {
  id      Int   @id @default(autoincrement())
  title   String
  userId  Int
  user    User  @relation(fields: [userId], references: [id])
}

medium

A. An error because 'include' is not valid here

B. An object with user data but 'posts' will be undefined

C. An object with user data and an array of two post objects under 'posts'

D. An array of posts only without user data

Solution

1. Step 1: Understand the Prisma query with include

   The query fetches a user by id and includes related posts as an array.

2. Step 2: Check relation and data

   User with id 1 has two posts, so 'posts' will be an array of two post objects inside the user object.

3. Final Answer:

   User object with posts array of two posts -> Option C

4. Quick Check:

   Include fetches related data as array [OK]

Hint: Include fetches related records as array or object [OK]

Common Mistakes:

• Expecting posts to be undefined without include
• Thinking include causes error
• Expecting only posts without user

4. Identify the error in this Prisma schema for a one-to-many relation between Author and Book models:

model Author {
  id    Int    @id @default(autoincrement())
  name  String
  books Book
}

model Book {
  id       Int    @id @default(autoincrement())
  title    String
  authorId Int
  author   Author @relation(fields: [authorId], references: [id])
}

medium

A. Author model is missing a primary key

B. Missing @relation on Author.books field

C. authorId should be optional in Book model

D. The 'books' field in Author should be an array (Book[]) not a single Book

Solution

1. Step 1: Check relation type from schema

   Author to Book is one-to-many, so Author.books must be an array of Book.

2. Step 2: Identify the mistake

   Author.books is declared as a single Book, which is incorrect for one-to-many.

3. Final Answer:

   'books' must be Book[] array -> Option D

4. Quick Check:

   One-to-many uses array field [OK]

Hint: One-to-many needs array [] on 'many' side field [OK]

Common Mistakes:

• Using single object instead of array for many side
• Forgetting @relation decorator
• Making foreign key optional incorrectly

5. You want to fetch all Order records with their related Customer data, but only include customers who have placed more than 5 orders. Which Prisma query approach correctly applies this filter using relations?

model Customer {
  id     Int     @id @default(autoincrement())
  name   String
  orders Order[]
}

model Order {
  id         Int      @id @default(autoincrement())
  total      Float
  customerId Int
  customer   Customer @relation(fields: [customerId], references: [id])
}

hard

A. Use findMany on Order with include: { customer: true } and a where on customer orders count > 5

B. Use findMany on Customer with where: { orders: { some: {} } } and include orders

C. Use findMany on Order with where: { customer: { orders: { count: { gt: 5 } } } } and include customer

D. Use findMany on Customer with include: { orders: true } and filter orders count > 5 in JavaScript

Solution

1. Step 1: Understand filtering by related model count

   Prisma allows filtering with nested conditions on related models using 'where' and relation filters.

2. Step 2: Identify correct query

   Using findMany on Order with include: { customer: true } and where: { customer: { orders: { _count: { gt: 5 } } } } fetches the desired orders with customers who have more than 5 orders.

3. Final Answer:

   findMany on Order with relational where filter and customer include -> Option A

4. Quick Check:

   Filter related model count in where clause [OK]

Hint: Filter related model count in where clause with include [OK]

Common Mistakes:

• Filtering in JavaScript instead of query
• Using incorrect nested where syntax
• Filtering on Customer without including orders count