Customizing the Lock Time-Out
When an instance of the Microsoft SQL Server Database Engine cannot grant a lock to a transaction because another transaction already owns a conflicting lock on the resource, the first transaction becomes blocked waiting for the existing lock to be released. By default, there is no mandatory time-out period and no way to test whether a resource is locked before locking it, except to attempt to access the data (and potentially get blocked indefinitely).
|In SQL Server 2005, use the sys.dm_os_waiting_tasks dynamic management view to determine whether a process is being blocked and who is blocking it. In earlier versions of SQL Server, use the sp_who system stored procedure.|
The LOCK_TIMEOUT setting allows an application to set a maximum time that a statement waits on a blocked resource. When a statement has waited longer than the LOCK_TIMEOUT setting, the blocked statement is canceled automatically, and error message 1222 (
Lock request time-out period exceeded) is returned to the application. Any transaction containing the statement, however, is not rolled back or canceled by SQL Server. Therefore, the application must have an error handler that can trap error message 1222. If an application does not trap the error, the application can proceed unaware that an individual statement within a transaction has been canceled, and errors can occur because statements later in the transaction might depend on the statement that was never executed.
Implementing an error handler that traps error message 1222 allows an application to handle the time-out situation and take remedial action, such as: automatically resubmitting the statement that was blocked or rolling back the entire transaction.
To determine the current LOCK_TIMEOUT setting, execute the @@LOCK_TIMEOUT function:
SELECT @@lock_timeout; GO