DB (De)Normalization

Normalization Organizing a DB to reduce redundancy by dividing tables into smaller related tables, establishing relationships using foreign keys. This requires queries with need for joins.

Denormalization Denormalization is process of combining related tames to reduce number of joins needed. This improves read performance by introducing some redundancy, but can lead to increased storage usage, and potential inconsistencies when updating data.

Normalized Schema

Customers Table

customer_id	name	email
1	Alice	alice@email.com
2	Bob	bob@email.com
Orders Table

order_id	customer_id	order_date	total_amount
101	1	2024-02-20	$50
102	2	2024-02-21	$30

• Data is stored efficiently
• Joins slow down read performance, especially with large tables

New Orders Table (Denormalized)

order_id	name	email	order_date	total_amount
101	Alice	alice@email.com	2024-02-20	$50
102	Bob	bob@email.com	2024-02-21	$30

• Faster reads (no join needed, direct query to one table)
• Redundancy (Customer details are duplicated in every order)
• Slower writes (if Alice changes her email, all rows need updating)

Impact Analysis

Operation	Before (Normalized)	After (Denormalized)
Read Performance	❌ Slower (requires JOINs)	✅ Faster (no JOINs)
Write Performance	✅ Faster (update in one table)	❌ Slower (update multiple rows)
Storage Efficiency	✅ Less redundancy	❌ More redundancy

Databases