dotnet/diagnostics
Diagnostics
The diagnostics project provides a cross-platform diagnostic infrastructure and command-line toolkit for monitoring runtime performance, analyzing memory dumps, and troubleshooting applications. It features a custom inter-process communication protocol for command and telemetry exchange across platforms, a low-overhead event pipe mechanism for streaming real-time diagnostic events and performance counters from running processes, and automated remote symbol resolution for stack trace analysis.
The platform includes native debugger extensions that integrate with standard debuggers to inspect managed code, threads, and internal runtime structures. It supports capturing lightweight managed heap memory graphs and process memory dumps by triggering garbage collection and object root stream capture. Additional capabilities include early source and symbol breakpoints, child process startup tracing, process discovery, and performance metric export in JSON or CSV formats.
Features
- .NET Command Line Interfaces - Provides a command-line toolkit and library collection for monitoring performance and troubleshooting applications.
- Debug Symbol Resolvers - Automatically downloads and resolves debugging symbols from remote symbol servers during stack trace analysis.
- Runtime Debugging - Extends native debuggers to inspect managed code, threads, and internal structures in active runtimes.
- Memory Dump Parsers - Captures and inspects process memory states using dedicated parsing utilities to diagnose runtime crashes.
- Java Heap Dump Capturers - Gathers lightweight managed heap memory graphs by triggering garbage collection and capturing object root streams.
- Debugger Extensions - Extends standard native debuggers with specialized plugins to inspect managed code and runtime structures.
- Runtime Diagnostics - Provides diagnostic tools for inspecting internal state, threads, and runtime structures of execution engines.
- Application Performance Monitoring - Tracks live performance counters and event streams from a running application to observe resource usage.
- Heap Snapshot Capturers - Captures managed heap memory snapshots and process dumps by triggering garbage collection and object root stream capture.
- Low-Overhead Runtime Event Streaming - Streams diagnostic events and performance counters in real time from running processes using a low-overhead circular buffer.
- Runtime Memory Explorers - Captures and inspects process memory dumps and live heap snapshots to diagnose memory leaks.
- Application Performance Profiling - Collects real-time diagnostic trace events and performance profiles to identify execution bottlenecks.
- Performance Tracing Platforms - Collects real-time events and generates performance traces from running processes to identify execution bottlenecks.
- Runtime Diagnostic Tracing - Captures diagnostic trace events from running applications across platforms without requiring a native profiler.
- Symbolic Breakpoints - Establishes breakpoints on managed methods or source file line numbers early in the debugging session.
- Debugging Symbol Managers - Connects to remote symbol servers or local directories to automatically download and resolve debugging symbols.
- Performance Counter Sampling - Collects live performance counter values and metrics from target processes for lightweight health monitoring.
- Inter-Process Communication - Communicates with target runtime processes across platforms using a custom inter-process communication protocol.
- Real-Time Application Performance Monitors - Collects real-time event traces from running applications to identify performance bottlenecks and track execution.
- Process Lifecycle Tracing - Launches application executables and immediately collects diagnostic traces from runtime startup.
- Performance Monitoring - Collects runtime performance counter values over time for lightweight health monitoring in production.
- Performance Metric Exports - Collects periodic counter data from a running process and exports it in JSON or CSV formats.
