How To Find Unused Indexes in SQL Server

dbtut April 23, 2019 MSSQL

Index is an object that improves performance in the database in general. Many inexperienced database administrators or application developers create too much index thinking that they always will improve performance. Generally, you do not need most of these indexes. Therefore, when a good database administrator starts looking at a database system, one of the first steps is to find and remove unused indexes in databases. You may also want to read the below articles.

“How To Find Missing Indexes in SQL Server“,

“How To Find Most Expensive Queries in SQL Server“,

“How To Find Unused Tables in SQL Server”

With the help of the following script, you can find unused indexes since the last restart of the sql server service.

select SCHEMA_NAME(tbl.schema_id) as [Schema Name],
OBJECT_NAME(us.object_id) as [Table Name],
	'DROP INDEX '+i.name + ' ON  [' + SCHEMA_NAME(tbl.schema_id)+'].['+ OBJECT_NAME(us.object_id)+ ']' as [Drop Index Script],
	us.last_user_seek [Last Seek Time],
	us.user_seeks [Seek Count],
	CASE us.user_seeks WHEN 0 THEN 0
		ELSE us.user_seeks*1.0 /(us.user_scans + us.user_seeks) * 100.0 END AS [Percent Of Seek],
	us.last_user_scan [Last Scan Time],
	us.user_scans [Scan Count],
	CASE us.user_scans WHEN 0 THEN 0
		ELSE us.user_scans*1.0 /(us.user_scans + us.user_seeks) * 100.0 END AS [Percent Of Scan],
	us.last_user_update [Last Update Time], --If the table is being used, this column will up to date. But this does not means the index is used.
	us.user_updates [Update Count]
FROM sys.dm_db_index_usage_stats us
INNER JOIN sys.indexes i ON i.object_id=us.object_id and i.index_id = us.index_id 
INNER JOIN sys.tables tbl ON tbl.object_id=us.object_id
WHERE us.database_id = DB_ID('Your_DB_Name') 
AND us.user_seeks=0 
AND us.user_scans=0 
AND i.is_primary_key = 0 -- To not return the Primary Key as a result. Even if it is not used, its presence is needed.
AND i.is_unique = 0      -- To not return the Unique Indexes as a result. Even if it is not used, its presence may be needed.
AND i.is_disabled=0      -- To not return the Disable Indexes as a result. Disable indexes can be deleted because they are not used.

select SCHEMA_NAME(tbl.schema_id) as [Schema Name],

OBJECT_NAME(us.object_id) as [Table Name],

'DROP INDEX '+i.name + ' ON [' + SCHEMA_NAME(tbl.schema_id)+'].['+ OBJECT_NAME(us.object_id)+ ']' as [Drop Index Script],

us.last_user_seek [Last Seek Time],

us.user_seeks [Seek Count],

CASE us.user_seeks WHEN 0 THEN 0

ELSE us.user_seeks*1.0 /(us.user_scans + us.user_seeks) * 100.0 END AS [Percent Of Seek],

us.last_user_scan [Last Scan Time],

us.user_scans [Scan Count],

CASE us.user_scans WHEN 0 THEN 0

ELSE us.user_scans*1.0 /(us.user_scans + us.user_seeks) * 100.0 END AS [Percent Of Scan],

us.last_user_update [Last Update Time], --If the table is being used, this column will up to date. But this does not means the index is used.

us.user_updates [Update Count]

FROM sys.dm_db_index_usage_stats us

INNER JOIN sys.indexes i ON i.object_id=us.object_id and i.index_id = us.index_id

INNER JOIN sys.tables tbl ON tbl.object_id=us.object_id

WHERE us.database_id = DB_ID('Your_DB_Name')

AND us.user_seeks=0

AND us.user_scans=0

AND i.is_primary_key = 0 -- To not return the Primary Key as a result. Even if it is not used, its presence is needed.

AND i.is_unique = 0 -- To not return the Unique Indexes as a result. Even if it is not used, its presence may be needed.

AND i.is_disabled=0 -- To not return the Disable Indexes as a result. Disable indexes can be deleted because they are not used.

Database Tutorials MSSQL, Oracle, PostgreSQL, MySQL, MariaDB, DB2, Sybase, Teradata, Big Data, NOSQL, MongoDB, Couchbase, Cassandra, Windows, Linux

How To Find Unused Indexes in SQL Server

About dbtut

Leave a Reply Cancel reply