Improvements were made to the way data was purged from the Data Repository database in SQL Monitor 2.2 and later. Purging is now performed as a faster, background task that reduces the affect on general performance of clearing a backlog of purged data. To facilitate this, SQL Monitor 2.2 and later has increased its usage of tempdb, so you may see tempdb grow by up to half the size of your Data Repository database when SQL Monitor is clearing a large backlog.

Once the backlog of data has been purged successfully, restart the SQL Server service and tempdb's size will revert back to normal.

A significant backlog of data that needs purging may build up if you perform any of the following actions:

• change your current purging policy to keep data in the Data Repository for a shorter amount of time
• remove a monitored server and delete all existing data for the server from the Data Repository

While the Data Repository database is busy with these actions, you may notice reduced performance while using the SQL Monitor interface.