In the previous article in this series, I showed you how data collector sets could free an administrator from having to select individual Performance Monitor Counters. Even so, knowing that a data collector set can make the selection easier is one thing, but actually performing an analysis is something entirely different. In this article, I will conclude the series by showing you how you can use the data collector sets.
Using Data Collector Sets
The first thing that you need to know about data collector sets is that although Microsoft has created them for you, they do not actually get used unless you tell Windows to run them. Fortunately, running a data collector set is easy. All you have to do is to right click on the data collector set that you are interested in, and then select the Start command from the resulting shortcut menu.
It is important to keep in mind that data collection is not an instantaneous process. Performance monitor counters are samples at various intervals, so you are going to need to let the data collector run for a while in order for you to get an accurate sampling of performance data. The actual length of time that the data collector set needs to run varies depending on what performance counters are involved, and on what you are trying to accomplish.
When you decide that you have collected an adequate sampling of data, then you must manually stop the data collector set. To do so, just right click on the data collector set, and choose the Stop command from the resulting shortcut menu.
It is worth pointing out that you really don’t have a lot of options when it comes to dealing with the default data collector sets. You can start or stop a data collector set, but that’s about it. The Reliability and Performance Monitor does give you the option of creating custom data collector sets though. When you create a custom data collector set, you have the option of building in a schedule so that the data collector set starts and stops at predetermined times. This can be extremely handy if you are trying to gather performance data corresponding to a particular time of the day.
Analyzing the Results
One of the things that makes data collector sets so nice is that you are not left having to figure out what all of the data means on your own. The Reliability and Performance Monitor console actually builds a report based on the data that has been collected.
Figure A This is an example of a data collector set report.
If you look at the console tree in the column above, you will notice that there is a Reports container located about halfway down the tree. If you expand this container as I have, you will find sub containers that correspond to the various data collector sets. Each of these sub containers contains all of the reports that have been run for the individual data collector set to date. For example, if you look at the System Performance container in the figure, you can see that there are currently two different reports that are archived on the system. You can view a report just by clicking on it.
One last thing that I want to mention about the reports is that they are displayed in a condensed format by default. If you look back at the figure, you will notice that the report is divided into sections, and that each section has an up or down arrow icon on the far right side of the section header. You can click the down arrow icons to get the console to show you more detail about a particular section. Clicking an up arrow icon condenses the section.
As you can see, data collector sets can greatly simplify the process of analyzing your system’s performance. Keep in mind that although Microsoft has not included a built in data collector set for every potential performance monitoring scenario, you always have the option of creating your own, custom data collector sets. To do so, just right click on the console’s User Defined container, choose the New | Data Collector Sets commands from the shortcut menus, and then fill in the blanks. It’s that easy.
Got a question? Post it on our Windows Server 2008 forums!