Treasure Hunt: Finding your application hot spot.
summary
Demonstrates concrete steps to find hot spots: use Visual Studio diagnostic tools and ETW sampling, compare sampling vs instrumentation, and run the Visual Studio profiler on a test machine. A WPF demo shows replacing create bitmap source with convert to bitmap source drops that call from 51% to 13% of CPU time; isolated apply filter stays as the real hot path. Deploying the same method to Azure and turning on Azure profiler enabled without code changes captures live traces and performance counters under load.
AI-written summary · about
chapters (18)
- 0:00Start
- 1:48Continuous Delivery
- 2:29Continuous Testing
- 3:38Performance Testing
- 4:50Three Steps for Perf Testing
- 5:43Simple Checklist
- 6:19General Guidelines
- 7:01Program Database Files (PDB)
- 11:04Build Configurations
- 13:13Observer Examples
- 13:40Debugging Performance
- 14:34Code Analysis
- 15:57Diagnostic Tools
- 37:00Sampling Profiler
- 37:27Instrumenting Profiler
- 41:53Monitoring Performance
- 42:53Application Insights
- 47:04Summary
related talks
01h 03mApplication Performance Optimisation in Practice
01h 17mThe Basics of Profiling
59mLessons Learned from the Trenches: Connectivity and Performance Tips&Tricks
38mExtending the Performance Analyis Toolset
01h 06mGet Better App Performance with this One Weird Trick
44m.NET Performance Investigation
01h 03mThe performance loop—A practical guide to profiling and benchmarking
33mProfile & Monitor Web Apps with Azure Application Insights