Creating Indexes in SQL – Syntax, Types, and Best Practices

Introduction

In SQL, text data types are used to store alphanumeric values like names, addresses, emails, and descriptions. Choosing the correct text type — CHAR, VARCHAR, or TEXT — is important for optimizing storage space, query speed, and database performance.

In this section, you'll learn the definitions, differences, and best use cases for each text data type.

1. CHAR (Fixed-Length String)

CHAR is used to store fixed-length strings. If the stored string is shorter than the defined length, SQL automatically pads it with spaces to match the specified size.

Features:

• Fixed length
• Fast and predictable performance
• Uses extra storage if the data is often shorter than the specified length

Syntax:

column_name CHAR(length);

length = number of characters (1 to 255 depending on the database system)

Example:

CREATE TABLE countries (
country_code CHAR(2),
country_name CHAR(50)
    );

country_code like 'US', 'IN', 'UK' will always take 2 characters.

When to Use CHAR:

• Data with a constant size, such as country codes, gender ('M', 'F'), state abbreviations
• Fixed-format fields like credit card types ('VISA', 'MC')
• When exact storage size is known and consistent

2. VARCHAR (Variable-Length String)

VARCHAR stands for Variable Character. It stores variable-length strings, meaning only the actual characters are stored without unnecessary padding.

Features:

• Variable length
• More space-efficient than CHAR for varying-length text
• Slightly slower than CHAR when processing large volumes (because of extra calculations for string lengths)

Syntax:

column_name VARCHAR(length);

length = maximum number of characters allowed

Example:

CREATE TABLE employees (
first_name VARCHAR(50),
email VARCHAR(100)
    );

Names and emails can vary in length, making VARCHAR ideal.

When to Use VARCHAR:

• Data with unpredictable or variable length
• Names, emails, addresses, and descriptions under 255-65535 characters
• Most general-purpose text fields

3. TEXT (Large Text Field)

TEXT is used to store large amounts of text like long descriptions, blog posts, comments, or articles.

Features:

• Meant for large text storage (up to 65,535 characters for standard TEXT in MySQL)
• Cannot have a default value (in some databases like MySQL)
• TEXT fields are stored outside the main table with a pointer reference
• Different variants exist (TINYTEXT, MEDIUMTEXT, LONGTEXT) for various sizes

Syntax:

column_name TEXT;

Example:

CREATE TABLE articles (
id INT,
title VARCHAR(255),
body TEXT
   );

body will store the full article content, which can be very large.

When to Use TEXT:

• Long-form text fields (comments, articles, reviews, reports)
• Data that exceeds normal VARCHAR limits
• When exact storage requirements are unknown or potentially very large

Quick Comparison: CHAR vs VARCHAR vs TEXT

Important Tips

• Use CHAR only when all values will be exactly the same length
• VARCHAR is the best choice for most standard text fields
• Reserve TEXT for content that exceeds VARCHAR limits
• Consider VARCHAR(MAX) in SQL Server for large text that might need indexing
• Be aware that TEXT fields may have limitations on default values and full-text indexing

What Is Index Creation in SQL?

Creating an index in SQL means instructing the database to build a data structure that speeds up search, filtering, and sorting operations on one or more columns. Proper index usage can dramatically improve query performance, especially on large datasets.

Basic Syntax for Creating an Index

CREATE INDEX index_name
ON table_name (column_name);

Example:

CREATE INDEX idx_lastname
ON employees (last_name);

This index helps accelerate queries like:

SELECT * FROM employees WHERE last_name = 'Smith';

Creating Index on Multiple Columns (Composite Index)

Composite indexes improve performance when queries filter or sort on two or more columns together.

CREATE INDEX idx_name_dept
ON employees (last_name, department_id);

This is useful for queries like:

SELECT * FROM employees
WHERE last_name = 'Doe' AND department_id = 3;

Order matters: (last_name, department_id) is not the same as (department_id, last_name).

Creating Unique Indexes

A unique index ensures that the indexed column (or combination) contains no duplicate values.

CREATE UNIQUE INDEX idx_email_unique
ON users (email);

Automatically enforces uniqueness, like an alternate to a UNIQUE constraint.

Index on Expressions (Supported in PostgreSQL and Others)

Some databases allow you to create indexes on expressions:

CREATE INDEX idx_lower_name
ON employees (LOWER(last_name));

This helps when searching case-insensitive:

SELECT * FROM employees
WHERE LOWER(last_name) = 'smith';

When to Create Indexes

Create indexes when:

• Columns are used frequently in WHERE, JOIN, ORDER BY, or GROUP BY
• You need to improve performance on large tables
• The column has many unique values

When Not to Create Indexes

Avoid over-indexing:

• On small tables (full table scans are faster)
• On columns with low cardinality (e.g., gender, status)
• On columns that change frequently (UPDATE, DELETE), as this slows write performance

Dropping an Index

Use this syntax to delete an index:

DROP INDEX index_name;

Some systems require table reference:

DROP INDEX index_name ON table_name;