I was recently asked by a friend, how can we control the level of locking in SQL Server when executing SELECT, UPDATE, INSERT and DELETE statements.
The answer to this question is by using locking hints. Locking hints direct SQL Server to the type of locks to be used. Even though the SQL Server Query Optimizer automatically determines the best locking option when executing a statement of the above kinds, there are cases where a DBA/Database Developer might needs to explicitly control the level of locking.
The available locking hints in SQL Server are the following:
HOLDLOCK
Applies to: SELECT, UPDATE, INSERT, DELETE
NOLOCK
Applies to: SELECT
PAGLOCK
Applies to: SELECT, UPDATE, INSERT, DELETE
READCOMMITTED
Applies to: SELECT, UPDATE, INSERT, DELETE
READPAST
Applies to: SELECT, UPDATE, DELETE
READUNCOMMITTED
Applies to: SELECT
REPEATABLEREAD
Applies to: SELECT, UPDATE, INSERT, DELETE
ROWLOCK
Applies to: SELECT, UPDATE, INSERT, DELETE
SERIALIZABLE
Applies to: SELECT, UPDATE, INSERT, DELETE
TABLOCK
Applies to: SELECT, UPDATE, INSERT, DELETE
TABLOCKX
Applies to: SELECT, UPDATE, INSERT, DELETE
UPDLOCK
Applies to: SELECT, UPDATE, INSERT, DELETE
XLOCK
Applies to: SELECT, UPDATE, INSERT, DELETE
The syntax for using locking hints is very simple. You only have to add the expression with (LOCKING_HINT) just right after you are referencing a database table name (or after a table alias) in your query.
An example of using the NOLOCK locking hint within a SELECT statement is the following:
SELECT column
FROM table with (NOLOCK)
The following MSDN Library link fully describes the above locking hints.
Labels: Locking Hints, SQL Server, SQL Server 2000, SQL Server 2005, SQL Server 2008