Why Keyset pagination for performance in Rest API? - Purpose & Use Cases

The Big Idea

What if you could skip the slow flipping and jump straight to the next page instantly?

The Scenario

Imagine you have a huge list of products in an online store. You want to show 10 products per page to your users. If you try to get page 100 by counting all previous products each time, it feels like flipping through a giant book page by page.

The Problem

Using simple page numbers means the system must count and skip many items before showing the right page. This makes the server slow and can cause delays or errors when many users browse deep pages.

The Solution

Keyset pagination uses a 'bookmark' from the last item seen to quickly jump to the next set. It avoids counting all previous items, making the system faster and smoother for users.

Before vs After

✗ Before

SELECT * FROM products ORDER BY id LIMIT 10 OFFSET 990;

✓ After

SELECT * FROM products WHERE id > :last_seen_id ORDER BY id LIMIT 10;

What It Enables

It enables fast, reliable browsing through large lists without slowing down or crashing the server.

Real Life Example

When scrolling through social media feeds or product catalogs, keyset pagination helps load new items instantly without waiting for the whole list to be counted.

Key Takeaways

Manual page counting slows down with big data.

Keyset pagination uses a last seen marker for speed.

This improves user experience and server performance.

Practice

1. What is the main advantage of using keyset pagination over traditional offset-based pagination in REST APIs?

easy

A. It requires less server memory by caching all pages in advance.

B. It allows jumping to any page number directly without sequential access.

C. It improves performance by avoiding slow offset queries on large datasets.

D. It automatically sorts data in descending order without extra code.

Solution

1. Step 1: Understand offset-based pagination issues

   Offset pagination uses a page number and offset, which becomes slow on large datasets because the database must skip many rows.

2. Step 2: Recognize keyset pagination benefits

   Keyset pagination uses a fixed key (like an ID) to fetch the next set of rows, avoiding the costly skip operation and improving performance.

3. Final Answer:

   It improves performance by avoiding slow offset queries on large datasets. -> Option C

4. Quick Check:

   Keyset pagination = better performance [OK]

Hint: Keyset uses keys, not offsets, for faster queries [OK]

Common Mistakes:

• Thinking keyset allows random page jumps
• Assuming keyset caches all data
• Believing keyset changes sort order automatically

2. Which of the following is the correct SQL snippet to implement keyset pagination for a table users ordered by id?

easy

A. SELECT * FROM users ORDER BY id LIMIT 10 OFFSET 20;

B. SELECT * FROM users WHERE id > 20 ORDER BY id LIMIT 10;

C. SELECT * FROM users WHERE id = 20 ORDER BY id LIMIT 10;

D. SELECT * FROM users ORDER BY id DESC LIMIT 10 OFFSET 20;

Solution

1. Step 1: Identify keyset pagination syntax

   Keyset pagination uses a WHERE clause with a key (like id > last_seen_id) and a LIMIT to fetch the next page.

2. Step 2: Analyze each option

   SELECT * FROM users WHERE id > 20 ORDER BY id LIMIT 10; uses WHERE id > 20 with ORDER BY id LIMIT 10, which matches keyset pagination logic.

3. Final Answer:

   SELECT * FROM users WHERE id > 20 ORDER BY id LIMIT 10; -> Option B

4. Quick Check:

   Keyset uses WHERE key > last_key [OK]

Hint: Keyset uses WHERE with key comparison, not OFFSET [OK]

Common Mistakes:

• Using OFFSET instead of WHERE for pagination
• Using equality (=) instead of greater than (>)
• Not ordering results by the key column

3. Given this REST API endpoint using keyset pagination:

GET /items?last_id=50&limit=5

And the database table items with IDs: [45, 47, 50, 52, 55, 60, 65], what will be the IDs returned by this request?

medium

A. [52, 55, 60, 65]

B. [45, 47, 50, 52, 55]

C. [52, 55, 60, 65, 70]

D. [51, 52, 53, 54, 55]

Solution

1. Step 1: Understand keyset pagination with last_id=50

   The API returns items with IDs greater than 50, limited to 5 results.

2. Step 2: Select IDs greater than 50 from the list

   IDs greater than 50 are [52, 55, 60, 65]. There are only 4 such items, so all are returned.

3. Final Answer:

   [52, 55, 60, 65] -> Option A

4. Quick Check:

   IDs > 50 limited to 5 = [52, 55, 60, 65] [OK]

Hint: Return items with ID > last_id, up to limit [OK]

Common Mistakes:

• Including items with ID equal to last_id
• Using offset instead of key comparison
• Assuming IDs are continuous numbers

4. You have this SQL query for keyset pagination:

SELECT * FROM orders WHERE order_date > '2024-01-01' ORDER BY order_date LIMIT 10;

But it returns duplicate rows when new orders are added. What is the likely cause?

medium

A. Ordering by a non-unique column causing duplicates

B. LIMIT value is too high causing repeats

C. Missing OFFSET clause to skip duplicates

D. Using > instead of >= in WHERE clause

Solution

1. Step 1: Identify ordering column uniqueness

   Ordering by order_date alone can cause duplicates if multiple rows share the same date.

2. Step 2: Understand keyset pagination requirements

   Keyset pagination requires ordering by a unique column or combination to avoid duplicates and missing rows.

3. Final Answer:

   Ordering by a non-unique column causing duplicates -> Option A

4. Quick Check:

   Order by unique key to avoid duplicates [OK]

Hint: Order by unique keys to prevent duplicates [OK]

Common Mistakes:

• Thinking OFFSET fixes duplicates
• Using >= instead of > causes duplicates
• Assuming LIMIT controls duplicates

5. You want to implement keyset pagination on a REST API that returns products sorted by price ascending, then by id ascending to break ties.
Which SQL WHERE clause correctly fetches the next page after last product with price=100 and id=50?

hard

A. WHERE price > 100 OR id > 50

B. WHERE price >= 100 AND id > 50

C. WHERE price > 100 AND id > 50

D. WHERE price > 100 OR (price = 100 AND id > 50)

Solution

1. Step 1: Understand multi-column keyset pagination

   When ordering by multiple columns, the WHERE clause must handle the first column and then the second to break ties.

2. Step 2: Analyze the correct condition

   The correct condition is price > 100 OR (price = 100 AND id > 50) to get all rows with higher price or same price but higher id.

3. Final Answer:

   WHERE price > 100 OR (price = 100 AND id > 50) -> Option D

4. Quick Check:

   Multi-column keyset uses OR + AND for tie-break [OK]

Hint: Use OR with AND to handle multi-column keys [OK]

Common Mistakes:

• Using AND instead of OR for first column
• Ignoring tie-break column in WHERE clause
• Using >= instead of > causing duplicates