Monitor System Resources

Every computer system has a performance threshold that, if pushed beyond this level, will cause the system to struggle to perform optimally. If you overload the system, it eventually slows down as it attempts to service each demand with the available resources. Most systems include a capable processor and sufficient amount of RAM for everyday or general needs. Memory is automatically reclaimed from apps that are closed. However, when apps or web browser tabs are left open, and more apps are then opened, the overall ability for the system to perform is degraded.

Understand baseline performance vs. real-time monitoring

You have seen that with tools such as Performance Monitor, Resource Monitor, and Task Manager, you can monitor your system activity and understand how demands on processor, RAM, networking, and disks affect your computer system. Real-time monitoring information is useful for instant diagnosis, whereas creating a baseline for your computer's performance can generate a system-specific report that can be useful to what your performance statistics look like during normal or heavy use.

If you intend to ship a device to a user who will use the device extensively for system-intensive tasks, such as video editing or computer-aided design, it might be useful to create a performance baseline for the device so that you can establish how the system performs normally and when under heavy load. This will be useful to confirm that the device specification is suitable for the user, but also if the user reports performance issues, you can run another performance baseline and compare the two baselines to evaluate whether the system environment has changed, for example, if the user regularly multitasks with additional new apps on the system that use additional memory.

In this scenario, when an issue or symptom occurs, you can compare your baseline statistics to your real-time statistic and identify differences between the two instances. When you can diagnose the issue, you can recommend a solution, such as to add more memory.

The most appropriate tool to record a baseline in Windows 10 is Performance Monitor; it will help you review and report on the following areas in your system.

• Evaluate your system workload.
• Monitor system resources.
• Notice changes and trends in resource use.
• Help diagnose problems.

Create a performance baseline

To create a performance baseline that monitors key system components you can use to measure against a future performance baseline, follow these steps

1. Click the Start button and type perfmon.
   Performance Monitor opens.
2. Click the Data Collector Sets node in the left pane.
3. Click User Defined, right-click User Defined, click New, and then click Data Collector Set.
4. In the Create New Data Collector Set Wizard, on the How Would You Like To Create This New Data Collector Set page, in the Name box, type Initial PC Baseline.
5. Click Create Manually (Advanced) and then click Next.
6. On the What Type Of Data Do You Want To Include page, select the Performance Counter check box and then click Next.
7. On the Which Performance Counters Would You Like To Log page, in the Sample Interval box, type 1 and then click Add.
8. Include the following counters.
   • Memory > Pages/Sec
   • Network Interface > Packets/Sec
   • PhysicalDisk > % Disk Time
   • PhysicalDisk > Avg. Disk Queue Length
   • Processor > % Processor Time
   • System > Processor Queue Length
9. Click OK and then click Finish.
10. Right-click Initial PC Baseline and then click Start.
11. Simulate load on the system by starting several programs, including Internet Explorer, Word 2016, Microsoft Excel 2016, and Microsoft PowerPoint 2016.
12. Close all Microsoft Office apps and Internet Explorer and stop the Initial PC Baseline data collector set.
13. To view the baseline report, in Performance Monitor, expand the Reports\User Defined node\Initial PC Baseline and click the report to open it.
14. Print the report or view the report and record the values for the following counters.
    • Memory > Pages/sec
    • Network Interface > Packets/Sec
    • PhysicalDisk > % Disk Time
    • PhysicalDisk > Avg. Disk Queue Length
    • Processor > % Processor Time
    • System > Processor Queue Length