Index selection guidelines in DBMS Theory - Step-by-Step Execution

Concept Flow - Index selection guidelines

Identify Query Patterns

↓

Check Column Usage in WHERE/JOIN

↓

Evaluate Column Selectivity

↓

Decide Index Type (Single/Multi-column)

↓

Create Index

↓

Monitor Query Performance

↓

Adjust or Drop Index if Needed

This flow shows how to choose columns for indexing by analyzing queries, checking column usage, evaluating selectivity, creating indexes, and monitoring performance.

Execution Sample

DBMS Theory

SELECT * FROM employees WHERE department_id = 5;
-- Index on department_id helps

SELECT * FROM employees WHERE last_name = 'Smith';
-- Index on last_name helps

SELECT * FROM employees WHERE salary > 50000;
-- Index may help if selective

These queries show how indexes on columns used in WHERE clauses can speed up data retrieval.

Analysis Table

Step	Action	Column Evaluated	Selectivity	Index Decision	Reason
1	Analyze query	department_id	High (few departments)	Create index	Used in WHERE, high selectivity
2	Analyze query	last_name	Medium (many unique names)	Create index	Used in WHERE, moderate selectivity
3	Analyze query	salary	Low (many salaries > 50000)	Consider index	Range query, less selective
4	Create index	department_id	N/A	Index created	Improves equality search
5	Create index	last_name	N/A	Index created	Improves equality search
6	Monitor queries	All	N/A	Adjust if needed	Check if indexes help performance
7	Drop index	salary	N/A	Index not created	Low benefit, maintenance cost high

💡 Index decisions made based on column usage and selectivity; low benefit indexes removed

State Tracker

Variable	Start	After Step 1	After Step 2	After Step 3	After Step 4	Final
department_id index	None	Considered	Considered	Considered	Created	Exists
last_name index	None	None	Considered	Considered	Created	Exists
salary index	None	None	None	Considered	Considered	Not created

Key Insights - 3 Insights

Why do we create an index on department_id but not on salary?

Because department_id has high selectivity (few departments), making the index efficient (see execution_table rows 1 and 3). Salary has low selectivity for the query, so the index is less useful and later dropped (rows 3 and 7).

What does 'selectivity' mean in index selection?

Selectivity means how many unique values a column has. High selectivity means many unique values, which makes indexes more effective (refer to execution_table column 'Selectivity').

Why monitor query performance after creating indexes?

Because indexes can improve or sometimes slow down queries. Monitoring helps decide if an index should be kept or dropped (see execution_table rows 6 and 7).

Visual Quiz - 3 Questions

Look at the execution_table, at which step is the index on last_name created?

A Step 5

B Step 2

C Step 4

D Step 7

💡 Hint

Check the 'Index Decision' column for last_name in execution_table rows

According to variable_tracker, what happens to the salary index after step 4?

A It is created then dropped

B It is created and remains

C It is never created

D It is created and monitored

💡 Hint

Look at the 'salary index' row in variable_tracker after step 4 and final

If department_id had low selectivity, what would likely change in the execution_table?

A Index would still be created

B Index decision would be to consider or drop

C No change in index decision

D Index would be created on salary instead

💡 Hint

Refer to how selectivity affects index decision in execution_table rows 1 and 3

Concept Snapshot

Index Selection Guidelines:
- Analyze queries to find columns used in WHERE/JOIN
- Evaluate column selectivity (unique values ratio)
- Create indexes on high-selectivity columns
- Use single or multi-column indexes as needed
- Monitor performance and adjust indexes
- Drop indexes with low benefit to save resources

Full Transcript

Index selection guidelines help decide which columns to index for faster queries. First, look at query patterns and find columns used in WHERE or JOIN clauses. Then check how selective these columns are—columns with many unique values are better for indexing. Create indexes on these columns to speed up searches. After creating indexes, watch query performance to see if they help. If an index does not improve performance or is costly to maintain, consider dropping it. This process ensures indexes improve database speed without wasting resources.