An index is a data structure associated with a table that provides fast access to rows in a table based on the values in one or more columns (the index key).

Let's say, you have a customers table in your database and you want to find out all the customers whose names begin with the letter A, using the following statement.


SELECT cust_id, cust_name, address FROM customers 
WHERE cust_name LIKE 'A%';

To find such customers, server must scan each row one by one in the customers table and inspect the contents of the name column. While it works fine for a table having few rows, but imagine how long it might take to answer the query if the table contains million of rows. In such situation you can speed things up by applying indexes to the table.

Creating an Index

You can create indexes with the CREATE INDEX statement:

CREATE INDEX index_name ON table_name (column_name);

For example, to create an index on the name column in the customers table, you could use:


CREATE INDEX cust_name_idx ON customers (cust_name);

Creating Multi-column Indexes

You can also build the indexes that span multiple columns. For example, suppose you've a table in your database named users having the columns first_name and last_name, and you frequently access the user's records using these columns then you can build an index on both the columns together to improve the performance, as follow:

CREATE INDEX user_name_idx ON users (first_name, last_name);

The Downside of Indexes

Index should be created with care. Because, every time a row is added, updated or removed from a table, all indexes on that table must be modified. Therefore, the more indexes you have, the more work the server needs to do, which finally leads to slower performance.

Here are some basic guidelines that you can follow while creating index:

  • Index columns that you frequently use to retrieve the data.
  • Don't create indexes for columns that you never use as retrieval keys.
  • Index columns that are used for joins to improve join performance.
  • Avoid columns that contain too many NULL values.

Also, small tables do not require indexes, because in the case of small tables, it is usually faster for the server to scan the table rather than look at the index first.

Drop Indexes

You can drop indexes that are no longer required with the following statement.

DROP INDEX index_name ON table_name;

The following statement will drop the index cust_name_idx from the customers table.


DROP INDEX cust_name_idx ON customers;