The Challenge of Intermittent Performance Issues
Troubleshooting Windows performance issues is easy when the problem is constant. If your computer is slow right now, you can open Task Manager, sort by CPU, and immediately identify the culprit. However, intermittent issues—like a server that spikes to 100% CPU usage every day at 3:00 AM, or a workstation that freezes randomly for five seconds while you are away from your desk—are incredibly difficult to diagnose because you cannot stare at Task Manager 24/7.
To solve this, Windows 11 includes a professional-grade diagnostic tool called Performance Monitor (PerfMon). PerfMon allows you to create Data Collector Sets that quietly log system metrics (like CPU, RAM, and Disk I/O) in the background over hours or days, allowing you to review the exact moment the spike occurred later.
Step-by-Step: Creating a Background CPU Log
Step 1: Open Performance Monitor
- Press the Windows Key, type
Performance Monitor, and hit Enter. - Ensure you are running it as an Administrator.
Step 2: Create a New Data Collector Set
- In the left-hand navigation pane, expand Data Collector Sets.
- Right-click on User Defined.
- Select New > Data Collector Set.
- Name the set “CPU Spike Tracker”.
- Select Create manually (Advanced) and click Next.
- Check the box for Performance counter under the “Create data logs” section, and click Next.
Step 3: Define the Metrics to Track
Now you must tell Windows exactly which sensors to record.
- Click the Add… button.
- In the list of available counters, locate and expand Processor.
- Select % Processor Time.
- In the box below labeled “Instances of selected object,” select _Total (to monitor the whole CPU) or select specific cores. Click the Add >> button.
- Optional but highly recommended: To figure out which app caused the spike, scroll down the available counters list, expand Process, select % Processor Time, select <All instances> in the bottom box, and click Add >>.
- Click OK to close the Add Counters window.
Step 4: Set the Polling Interval
Back in the setup wizard, you will see a Sample interval setting. This defines how often Windows records the data.
- Setting it to 1 second gives incredibly detailed data but creates massive log files.
- For monitoring over a 24-hour period, setting the interval to 15 Seconds is usually the sweet spot between capturing spikes and keeping the file size manageable.
Click Finish to save the Collector Set.
Step 5: Start the Logging Process
- In the left pane, click on User Defined to view your new “CPU Spike Tracker”.
- Right-click it and select Start.
A small green play icon will appear next to it. You can now close Performance Monitor. Windows is quietly writing your CPU metrics to a log file in the background. Leave the computer running until the intermittent issue occurs.
Reviewing the Data
Once the issue has occurred, you must stop the logging and analyze the results.
- Open Performance Monitor again.
- Navigate to Data Collector Sets > User Defined, right-click “CPU Spike Tracker”, and select Stop.
- To view the data, look at the left pane, expand Reports > User Defined > CPU Spike Tracker.
- Click on the newly generated report file.
A graph will appear showing the CPU usage over the entire recording period. You can visually identify the massive spikes on the timeline. If you added the Process counters in Step 3, you can highlight that specific timestamp and scroll through the list below the graph to see exactly which application (e.g., chrome.exe or MsMpEng.exe) was consuming the CPU at that exact moment.
Conclusion
Task Manager is a tool for the present moment, but Performance Monitor is a tool for history. By configuring a Data Collector Set, IT professionals can eliminate guesswork and rely on hard data to identify the root cause of intermittent Windows performance degradation.