Skip to content

Instantly share code, notes, and snippets.

@nolanlawson
Last active November 15, 2019 23:47
Show Gist options
  • Star 4 You must be signed in to star a gist
  • Fork 2 You must be signed in to fork a gist
  • Save nolanlawson/f5244175fa06121b4eacc0e8c5206db7 to your computer and use it in GitHub Desktop.
Save nolanlawson/f5244175fa06121b4eacc0e8c5206db7 to your computer and use it in GitHub Desktop.
Instructions for taking an ETL trace for Edge or IE

How to take a trace of a website issue in Edge/IE

Taking a trace

Method 1: Feedback Hub

  1. In Feedback Hub, click "Add new Feedback"
  2. Choose "Microsoft Edge" as the category
  3. Choose "Website issues" as the sub-category
  4. Click "Recreate my problem"
  5. Ensure that both checkboxes are checked ("include data" and "include screenshots")
  6. Click "Start capture"
  7. Reproduce the issue
  8. Click "Stop capture" in Feedback Hub
  9. Click "Submit and upvote"

Method 2: Using Windows Performance Toolkit

  1. Download the Windows Performance Toolkit (WPT) via the Windows ADK: http://aka.ms/adk
  2. Close Edge/IE completely (in case you had F12 open, which can change the performance characteristics)
  3. Open Edge/IE
  4. Navigate to the URL about:blank
  5. Run Windows Performance Recorder (WPRUI.exe)
  6. Click "More options"
  7. Enable the following options:
  • First-level triage
  • CPU usage
  • Edge Browser (or "Internet Explorer" for IE)
  • HTML Responsiveness Analysis
  1. Change "Logging mode" from "Memory" to "File" (tends to preserve ETW events better)
  2. Click "Start" to start recording
  3. Reproduce the scenario (page load, website interaction, etc.)
  4. Click "Save" to save the ETL file
  5. Zip the ETL file to save space

Analyzing a trace

  1. Open the ETL or the zipped ETL file in Windows Performance Analyzer (wpa.exe)
  2. Click Trace -> Load Symbols
  3. Click Profiles -> Apply -> Browse Catalog -> HtmlResponsivenessAnalysis.wpaProfile
  4. Use the "Frame Analysis" tab to analyze CPU usage
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment