N+1 query problem in Spring Boot - Performance & Optimization

Performance: N+1 query problem

HIGH IMPACT

This affects page load speed by causing many database queries during data fetching, which delays server response and slows down rendering.

Fetching a list of entities with their related data

Spring Boot

List<Order> orders = orderRepository.findAllWithCustomers(); // uses JOIN FETCH or EntityGraph
// customers loaded in one query with orders

Fetches orders and customers in a single query, reducing database calls drastically.

📈 Performance GainSingle query regardless of N, cutting database calls from N+1 to 1.

Fetching a list of entities with their related data

Spring Boot

List<Order> orders = orderRepository.findAll();
for (Order order : orders) {
  Customer customer = customerRepository.findById(order.getCustomerId()).orElse(null);
  order.setCustomer(customer);
}

This runs 1 query to get orders plus N queries to get each customer, causing many database hits.

📉 Performance CostTriggers N+1 database queries, increasing server response time linearly with N.

Performance Comparison

Pattern	DOM Operations	Reflows	Paint Cost	Verdict
N+1 Query Pattern	No direct DOM impact	No direct reflows	Blocks initial paint due to slow server response	[X] Bad
Batch Fetching with JOIN FETCH	No direct DOM impact	No direct reflows	Faster initial paint due to quicker server response	[OK] Good

Rendering Pipeline

The server delays sending HTML until all data queries finish. Multiple queries increase server processing time, delaying HTML delivery and first paint.

→Server Data Fetching

→HTML Generation

→First Paint

⚠️ BottleneckServer Data Fetching due to multiple database queries

Core Web Vital Affected

LCP

This affects page load speed by causing many database queries during data fetching, which delays server response and slows down rendering.

Optimization Tips

1Avoid running one database query per item in a list; batch fetch related data.

2Use JOIN FETCH or EntityGraph annotations in Spring Boot to load related entities eagerly.

3Reducing database queries improves server response time and speeds up page load (LCP).

Performance Quiz - 3 Questions

What is the main performance problem caused by the N+1 query pattern?

A Multiple database queries increase server response time

B Too many DOM nodes are created

C CSS selectors become very complex

D JavaScript bundle size increases

DevTools: Network

How to check: Open DevTools, go to Network tab, reload page, and observe the number and timing of API/database calls or backend requests.

What to look for: Look for many sequential API calls or slow backend responses indicating multiple queries causing delays.

Practice

1. What is the N+1 query problem in Spring Boot applications?

easy

A. Not using any database queries at all

B. Making only one query to fetch all data including related entities

C. Using incorrect SQL syntax in queries

D. Making one query to fetch a list, then one query per item to fetch related data

Solution

1. Step 1: Understand the query pattern

   The N+1 problem occurs when the app first fetches a list (1 query), then fetches related data for each item separately (N queries).

2. Step 2: Identify the problem impact

   This causes many queries, slowing down the app and wasting resources.

3. Final Answer:

   Making one query to fetch a list, then one query per item to fetch related data -> Option D

4. Quick Check:

   N+1 query problem = multiple queries instead of one [OK]

Hint: N+1 means 1 query + N queries for related data [OK]

Common Mistakes:

• Thinking N+1 means only one query is made
• Confusing it with syntax errors
• Assuming it is about missing queries

2. Which of the following is the correct way to use JOIN FETCH in a Spring Data JPA query to avoid the N+1 problem?

easy

A. @Query("SELECT o FROM Order o JOIN FETCH o.items")

B. @Query("SELECT o FROM Order o JOIN o.items")

C. @Query("SELECT o FROM Order o LEFT JOIN o.items")

D. @Query("SELECT o FROM Order o WHERE o.items IS NOT NULL")

Solution

1. Step 1: Understand JOIN FETCH usage

   JOIN FETCH tells JPA to fetch related entities eagerly in one query, avoiding multiple queries.

2. Step 2: Identify correct syntax

   @Query("SELECT o FROM Order o JOIN FETCH o.items") uses JOIN FETCH correctly to fetch orders with their items in one query.

3. Final Answer:

   @Query("SELECT o FROM Order o JOIN FETCH o.items") -> Option A

4. Quick Check:

   JOIN FETCH = eager fetch to avoid N+1 [OK]

Hint: Use JOIN FETCH to load related data in one query [OK]

Common Mistakes:

• Using JOIN without FETCH causes lazy loading
• Using WHERE instead of JOIN FETCH
• Missing FETCH keyword

3. Given this Spring Data JPA repository method:

@Query("SELECT c FROM Customer c")
List<Customer> findAllCustomers();

And assuming Customer has a lazy-loaded orders collection, what happens when you call findAllCustomers() and then access orders for each customer?

medium

A. One query to get customers, then one query per customer to get orders (N+1 problem)

B. One query to get customers and all orders in one go

C. No queries are made until orders are accessed

D. An error occurs because orders are not fetched

Solution

1. Step 1: Analyze the query and lazy loading

   The query fetches customers only; orders are lazy-loaded, so not fetched initially.

2. Step 2: Accessing orders triggers queries

   Accessing orders for each customer triggers one query per customer, causing N+1 queries total.

3. Final Answer:

   One query to get customers, then one query per customer to get orders (N+1 problem) -> Option A

4. Quick Check:

   Lazy loading causes N+1 queries [OK]

Hint: Lazy loading causes one query per item when accessed [OK]

Common Mistakes:

• Assuming all data loads in one query
• Thinking no queries run until orders accessed
• Confusing lazy and eager loading

4. You have this code snippet causing N+1 queries:

List<Author> authors = authorRepository.findAll();
for (Author a : authors) {
    System.out.println(a.getBooks().size());
}

How can you fix it to avoid the N+1 problem?

medium

A. Add @Transactional annotation to the method

B. Call getBooks() inside a separate thread

C. Change repository method to use @Query("SELECT a FROM Author a JOIN FETCH a.books")

D. Remove the loop and print authors only

Solution

1. Step 1: Identify cause of N+1

   Calling getBooks() inside loop triggers one query per author due to lazy loading.

2. Step 2: Use JOIN FETCH to load books eagerly

   Changing repository query to use JOIN FETCH loads authors and books in one query, avoiding N+1.

3. Final Answer:

   Change repository method to use @Query("SELECT a FROM Author a JOIN FETCH a.books") -> Option C

4. Quick Check:

   JOIN FETCH fixes N+1 by eager loading [OK]

Hint: Use JOIN FETCH in query to load related data eagerly [OK]

Common Mistakes:

• Adding @Transactional does not fix N+1
• Using threads does not solve query count
• Removing loop hides problem but does not fix it

5. You have entities Post and Comment with a one-to-many lazy relationship. You want to fetch all posts with their comments efficiently. Which approach best avoids the N+1 problem and handles posts with no comments?

hard

A. Use native SQL without JOIN FETCH and map manually

B. @Query("SELECT p FROM Post p LEFT JOIN FETCH p.comments") to fetch posts and comments in one query

C. Fetch posts first, then fetch comments in a separate query for each post

D. Fetch posts only and ignore comments to reduce queries

Solution

1. Step 1: Understand lazy loading and N+1

   Lazy loading comments causes one query per post when accessed, causing N+1 problem.

2. Step 2: Use LEFT JOIN FETCH to include posts without comments

   LEFT JOIN FETCH fetches posts and their comments in one query, including posts with no comments.

3. Final Answer:

   @Query("SELECT p FROM Post p LEFT JOIN FETCH p.comments") to fetch posts and comments in one query -> Option B

4. Quick Check:

   LEFT JOIN FETCH avoids N+1 and includes empty collections [OK]

Hint: Use LEFT JOIN FETCH to include all posts and comments [OK]

Common Mistakes:

• Using INNER JOIN FETCH excludes posts without comments
• Fetching comments separately causes N+1
• Ignoring comments loses needed data