SQL Foundations  «Prev  Next»

Lesson 5What is an index?
ObjectiveUse Database Table Index to speed up your Queries.

Use Table Index to speed up Queries

Using indexes in your SQL database can significantly increase the performance of your queries by enabling faster data retrieval.
Here's a guide on how you can utilize database table indexes to speed up your SQL queries:
  1. Understand the Basics of Indexing: A database index works much like an index in a book; it allows the database to find and retrieve specific rows much faster than it could without an index. But unlike a book's index, the database maintains its indexes automatically. Once an index is created, the database system manages it and uses it whenever it can.
  2. Create Indexes Based on Your Queries: Indexes are most effective when they're designed with your queries in mind. Analyze the queries that are running slow and identify the columns that are used in WHERE clauses, JOIN conditions, ORDER BY, GROUP BY, or SELECT statements. Those are your prime candidates for indexing.
    CREATE INDEX idx_product_description 
    ON Products (Product_Description);
    

    This statement creates an index named 'idx_product_description' on the 'Product_Description' column in the 'Products' table.
  3. Use Composite Indexes for Complex Queries: If your queries often filter or sort on multiple columns together, consider creating a composite index, which includes multiple columns. The order of columns in the index definition can matter for these multi-column indexes, generally aligning the column order in the index with your query patterns can lead to better performance.
    CREATE INDEX idx_product 
    ON Products (Product_ID, Product_Description);
    
  4. Limit the Number of Indexes: While indexes can speed up query performance, they come with a cost. Every time you modify data (insert, update, or delete), your indexes need to be updated as well. Having too many indexes can slow down these operations and take up additional storage space. Balance is key.
  5. Monitor and Optimize Your Indexes: Regularly monitor the performance of your indexes and adjust as necessary. You may find that some indexes are not used often, and they could be dropped. Database management systems often include tools for index monitoring and reporting.
  6. Keep Your Database Statistics Up to Date: Database systems use statistics about the distribution of data in your table to make decisions about when and how to use indexes. Keeping these statistics up to date can help the query planner make better decisions and potentially improve performance.

Remember, while indexes can improve database performance, they are not a magic bullet. Always measure the performance of your queries before and after adding indexes to make sure they're having the desired effect.
As you start working with SQL in your database, you will find that you will pay more and more attention to making your query run faster and return results more quickly. Although you will be able to accomplish some of this simply by working with the query itself (you can remove unneeded columns, decrease the number of rows that will be returned, and so on), it is also likely that an index will help.

Table Index

An index is a document that is manually created by the database. It helps the database find information more quickly by indexing locations.
To help understand what an index is about, you might want to consider your favorite file cabinet. If the folders are in the cabinet in random order, finding a folder in short order will be next to impossible. It is more likely that you sort the folders, perhaps adding dividers to help you quickly go to the section of folders that contain the specific folder you need. With a SQL database, the concept is the same. You will need to decide how you are going to be asking for information in the database, and then use this information to determine if it makes sense to create an index that directly supports that approach. How to create an index will be discussed next.

SQL Queries

Indexes and SQL Tables

Question: I have often heard the term indexes discussed within the context of creating SQL tables. How do you create indexes?
Answer:
Creating indexes in SQL involves using specific SQL commands, typically `CREATE INDEX`. Here's how you can create indexes in various SQL databases:
Basic Syntax for Creating an Index
CREATE INDEX index_name
ON table_name (column1, column2, ...);

Detailed Steps and Considerations:
  1. Choosing the Index Name: You can name your index anything, but it's good practice to make it descriptive, e.g., `idx_user_email` for an index on the email column in a users table.
  2. Selecting the Table: Specify the table on which you want to create the index.
  3. Specifying Columns: List the columns you want to index. Multiple columns can be specified for composite indexes.

Creating a Single Column Index:
CREATE INDEX idx_user_email ON users(email);

This creates an index named `idx_user_email` on the `email` column in the `users` table.
Creating a Composite Index:
CREATE INDEX idx_user_name_birthday ON users(last_name, first_name, birthday);

This creates a composite index on `last_name`, `first_name`, and `birthday` in the `users` table, which would be useful for queries involving these fields together.
Types of Indexes:
  • Unique Index: Ensures that the indexed columns contain unique values.
    CREATE UNIQUE INDEX idx_unique_email ON users(email);
    
  • Clustered Index: Physically reorders the table to match the index order (SQL Server specific, similar to the primary key in other systems).
    CREATE CLUSTERED INDEX idx_clustered_id ON users(id);
    
  • Non-Clustered Index: Does not alter the physical order of the table (default in most systems).
    CREATE NONCLUSTERED INDEX idx_nonclustered_name ON users(last_name);
    

Additional Options:
  • Include Columns** (SQL Server): Allows you to include non-key columns in the leaf level of the index to cover more queries.
    CREATE INDEX idx_with_include ON orders(order_id) INCLUDE (customer_id, order_date);
    
  • Index on Expressions: Some databases allow creating indexes on expressions or functions.
    CREATE INDEX idx_upper_name ON users(UPPER(last_name));
    

Considerations:
  • Performance vs. Storage: Each index increases the size of the database and can slow down write operations. Balance the need for read performance with the overhead.
  • Maintenance: Indexes need to be maintained. Consider using `FILLFACTOR` in SQL Server or similar options in other databases to optimize index maintenance.
  • Statistics: Ensure that the database's statistics are up-to-date, as these are used by the query optimizer to decide when to use an index.
  • Indexing Strategy: Not all columns need indexes. Typically, columns used in `WHERE`, `JOIN`, `ORDER BY`, or `GROUP BY` clauses benefit from indexing.
Conclusion:
Creating indexes is about optimizing your database for the queries you run most often. It's an art as much as a science, requiring an understanding of your data access patterns. Always test the performance impact of new indexes in a development or staging environment before applying them to production.

SEMrush Software