question

I first noticed the problem with this

Configuation_No1 Server1: SQL Server 2008 database I am tracing Workstation: my computer where I have setup client side trace Server2: location where I am writing the trace to a SQL Server 2008 table. Template: Standard Default

Results: I confirmed that while Profiler is runnning one of my applications has a major slowdown.

Configuration_No2 Server1: SQL Server 2008 database I am tracing Server1: Server Side Tracing Server1: Server Side Saving Trace to Table Template: Standard Default

Results: This configuration also causes a slowdown.

You should not be using Profiler to monitor your server. The Profiler GUI is primarily for browsing data after it's been collected, not for active monitoring. You should set up a server side trace using the trace events and output the results to a file. You can then use Profiler to browse the data from the file after it's completed. If you're unsure how to do this, set up the trace you'd like to do using the Profiler GUI and then use the Export function to save it as a trace definition.

The reason I say this is because a server-side trace and the Profiler actually collect data in totally different ways. The Profiler can cause serious performance problems on the server where server side trace events don't. I've also seen Profiler drop data that the trace events capture.

Client connections locking by profiler happens to us too. Note its not a major slow down or anything - it's a complete halt of all connections that Finance application has to the database server. All existing instances freeze, and you can't open a new one. Happens the instance profiler trace is turned on, and gets fixed as soon as its turned of. SQL 2008 R2, on Windows Server 2008. Max number of connections set to 0. The application can be used by 50 people across our WAN simultaneously no problem. It's just the profiler that's causing it.