Client Application Performance Testing: Using Process Explorer and Windows Performance Toolkit

When a client application shows performance issues, simply opening Task Manager only gives a macro view of overall CPU, memory, disk, and network usage. To inspect specific threads or micro‑level CPU consumption, more specialized tools are required.

Validation Method 1 – Process Explorer

Use Process Explorer to locate the main process (e.g., 360rp.exe ), open its Properties , and view thread details. Identify the target thread (e.g., CleanupAllClientRules ) and examine its CPU and Cycles Delta values, which indicate the amount of CPU work per unit time.

Validation Method 2 – Windows Performance Toolkit (WPT)

WPT, included in the Windows SDK, provides two main tools: Windows Performance Recorder (WPR) for capturing system events and Windows Performance Analyzer (WPA) for analyzing the generated .etl logs.

WPT Installation

1. Download the Windows SDK from https://developer.microsoft.com/zh-cn/windows/downloads/sdk-archive/ . 2. Run winsdksetup.exe and follow the default installation steps, ensuring the "Windows Performance Toolkit" component is selected. 3. After installation, the tools are located at C:\Program Files (x86)\Windows Kits\10\Windows Performance Toolkit\ .

Using WPR – Recording a Log

Launch WPRUI.exe , select the desired metrics (e.g., CPU usage, File I/O, Heap usage), configure the scenario (e.g., Boot, Verbose), start recording, and later click Save to generate an .etl file.

Using WPA – Analyzing the Log

Open the .etl file in WPA, filter by "Utilization By Process and Thread*", and select the specific process. Apply "Filter to Selection" to isolate the process, then examine the thread list to find the target thread (e.g., ID 5744). Compare the Cycles and Cycles Delta values before and after code changes; a significant reduction indicates successful optimization.

Conclusion

Both Process Explorer and WPT can reveal micro‑level CPU usage that Task Manager cannot. Understanding each tool's strengths enables more effective performance testing, and integrating such tests into CI pipelines moves performance verification toward automation.