Application Diagnostics in .NET Core 3.1
summary
Shows concrete .NET Core 3.1 diagnostics: events/metrics/dumps as primitives, EventPipe as an always‑on named pipe, and live APIs like DiagnosticSource/Activity for distributed traces. Demonstrations use dotnet-trace, dotnet-counters and dotnet-dump to collect CPU profiles, counters, mini/GC dumps and open them in Visual Studio. A GC dump example is tiny (~1K) vs bulky full dumps, a pathological app shows ~750 threads and blocking, and the new allocation profiler plus parallel tasks view expose hot types and waiting logical stacks.
AI-written summary · about
chapters (10)
- 0:00Introduction to diagnostics
- 0:53Diagnostics primitives
- 3:51Diagnostic infrastructure
- 7:11Implementing diagnostics
- 11:55Collecting diagnostic data
- 16:44CLI tools and visualization
- 25:06Distributed tracing
- 39:38Memory leak analysis
- 53:35Debugging async issues
- 59:40Crash reporting and wrap-up
related talks
01h 01mBeyond step-by step debugging in Visual Studio
01h 04mWeb Diagnostics with a Glimpse into ASP.NET
01h 06mGet Better App Performance with this One Weird Trick
38mLogging, Instrumentation, Dashboards, Alerts
41mModern Application Debugging: An Introduction To OpenTelemetry
57mDon’t need to understand the details of .NET. Until you do
43m.Net Standard & .Net Core: Platform Evolution
06mon their Application Insights tools