Report Description: Profiler Trace Analysis

The purpose of the Profiler Trace Analysis top-level report is to graph the rate of query execution and the amount of CPU and I/O used over time, and to identify the most expensive queries in your workload. 

 

Use the graph of query execution statistics at the top of the report to identify which type of activity appears to be the bottleneck in your system.  Then use the Expensive Queries list at the bottom of the report to identify the query that is using the most of that resource.  For example, suppose that you had imported a profiler trace that was captured when your server was CPU-bound.  You would sort the Expensive Queries list by CPU use by clicking the sort icon on the CPU column header (click the icon twice to sort in descending order).  Once you have identified the offending query, click the query text to drill into a similar report showing the execution statistics for that particular query.  In the child report you can see whether the rate of execution of the query increased, or whether the CPU cost per execution changed, for example due to a different query plan. 

 

Other comments: