'How To' Series (Part 1): Performance Monitor
September 5, 2001
I am going to run a brief series of articles stepping through some of the more common administrative tools that a Windows 2000 administrator might have to use. Although some of the articles might seem trivial to you, it's probably only because you have high exposure to these tools on a day-to-day basis. For some of the younger folks and newer admins, this information should prove quite useful. And, for you older folks, maybe you'll pick up a thing or two.
So, the first tool I am going to discuss is one of the most fundamental tools that we use in any Server (or workstation for that matter) troubleshooting situation. This tool is the Performance Monitor.
The Performance Monitor is accessed in Windows 2000 though the Control Panel's Administrative Tools icon. When you click on "Performance" you will be presented with the MMC for the Performance Monitor. It looks like the following:
To do any work within this tool, you must be in the right side of the window, the active side. If you want to add counters to the live view of your machine, simply right click and select "Add Counters". You will be presented with the following:
From here, you can select what counters you would like to view and from where. Be careful not to add too many, or viewing real time will prove cumbersome.
You can also build a template for monitoring machines in your environment via this tool. If you collect a set of Counters that you would like to set as your standard, you can right click, drop to Save As, and save these as an HTM template. With NT, you would be saving them as a PML or something of that format.
To collect logging information from that template or just to monitor logging of Performance data at all, you would select Performance Logs and Alerts from the left side of your screen. There is an example of a performance monitor log that you can start and stop to see the effects. You can have this log information piped to a BLG file, a CSV, or basically any format that you would like to collect in.
You can also schedule the stopping and starting of logging by dropping to "Properties" of a log. Here, you will be able to input a stop and start time. This can be handy if you are having recurring problems on a server during a weekend rollout or something like that. You would not want to collect data from 48 hours if you only need 12 hours worth...
You can import and view data from these logs as if you were viewing your own active information (only it will be a static view). You can actually pick and choose which counters you want to view from the logged data the same way you would choose from active performance views (as described above).
I hope this information helps some of you along. Let me know if you have a tool that you would like me to explain in an article.