A latch is a "lightweight lock". A latch acts like a lock, in that its purpose is to prevent data from changing unexpectedly. For example, when a row of data is being moved from the buffer to the SQL Server storage engine, a latch is used by SQL Server during this move to prevent the data in the row from being changed during this very short time period. This not only applies to rows of data, but to index information as well, as it is retrieved by SQL Server.

Just like a lock, a latch can prevent SQL Server from accessing rows in a database, which can hurt performance. Because of this, latch wait time must be minimized.

The MsSqlLatchTest makes the following measures of latch activity for an Microsoft SQL database.

 

Measurement	Description	Measurement Unit	Interpretation
Latch_wait_rate	This is the number of latch requests per second that could not be granted immediately.	Waits/Sec	The latch activity may vary from one server to another. Hence, it is important to get baseline numbers for this metric, so that you can compare "typical" latch activity against what is happening currently. If latch activity is higher than expected, this often indicates one of two problems. First, it may mean that the SQL Server could use more memory. If latch activity is high, check the buffer cache hit ratio is. If it is below 99%, the SQL server could probably benefit from more memory. If the hit ratio is above 99%, then it could be the I/O system that is contributing to the problem, and a faster I/O system might improve server's performance.
Total_wait_time	This metric denotes the total wait time for latch requests that had to wait in the last second.	Secs	Ideally, this value should be close to 0. The larger this value is, the more contention there is for latches and worse the performance of the database. If the wait time is high, check the SYSPROCESSES table of the database to see which latches are seeing most contention.