

Right next to this CPU summary rectangle is a breakdown of current CPU utilization by resource pool: One thing to note here is that all of these numbers are pulled from the ring buffers, and this data is only refreshed once a minute (for more information on this please see Amit Banerjee’s blog post on the topic ). So this data can be up to 60 seconds stale, and if your CPU utilization for each resource pool (real-time) doesn’t make sense compared to this CPU summary rectangle (refreshed every 60 seconds) then it is because of this update schedule for the ring buffers. Then right below this there is a little sparkline graph to visually show the past hour of this data. This rectangle has a bit of information, including three CPU metrics: Total System CPU utilization, SQL Server process CPU utilization, and other process CPU utilization. On the left hand side you can see the system CPU summary: This is the CPU monitoring portion of the report. In the main screenshot, focus on the top row of data. The all-encompassing screenshot can be seen below: I designed the report to have an overview breakdown of CPU and memory utilization, and then by each particular resource pool in their respective category. As SQL Server DBAs, SSMS is our go-to tool of choice. I have written an SSRS report (well, actually two reports, but logically it can be considered one monitoring solution) to show this data. The reason I chose to write an SSRS report is because I wanted the ability to bring it into SQL Server Management Studio as a custom report. The challenge I’m looking to tackle here today is a solution in order to visualize what is currently happening with resource consumption. What I will be talking about today is a tool, in particular a report, that I’ve developed to monitor the current state of Resource Governor usage and the current activity that is taking place. I will not be going into the details of Resource Governor, or even how to set it up as this is extensively documented in Books Online.

Resource Governor is an extremely cool and powerful feature that, when implemented, can manage the consumption of CPU and memory resources.
