Exploring your .NET applications with dotnet-monitor

I talked about Cross-platform diagnostic tools for .NET Core and dotnet-trace for .NET Core tracing but I would be remiss if I didn't show and mention "dotnet monitor."

dotnet monitor is an experimental tool that makes it easier to get access to diagnostics information in a dotnet process. If you're running .NET Core within a container - or mostly likely in Kubernetes - this tool offers an insight into your running microservice. Basically it creates a microservice of its own that you can interrogate to better understand what's happening.

HELP! Confused about Kubernetes? Check out http://computerstufftheydidntteachyou.com and my most recent video on Kubernetes and Container Orchestration.

Assuming you have .NET Core installed, you can install dotnet monitor as a global tool:

dotnet tool install -g dotnet-monitor --add-source https://dnceng.pkgs.visualstudio.com/public/_packaging/dotnet5-transport/nuget/v3/index.json --version 5.0.0-preview.*

You then just run it along side your project or running process.

dotnet monitor collect

The developer blog on dotnet monitor shows you how you can share a volume mount between your application container and a container running dotnet monitor if you like. If you're in k8s (Kubernetes) you should run dotnet monitor as a sidecar to your container within the same pod.

It'll start up and you can talk to dotnet monitor with curl, wget, and pipe through jq and hit localhost:52323/processes to get a list of .NET Core processes it can think about.

NOTE: If you are running this locally and get auto redirected to HTTPS then you may have a cached HSTS policy for localhost from other work. Head over to edge://net-internals/#hsts (or chrome://) and scroll to the bottom and delete the Domain Security Policies for localhost.

Now I can curl and see the output. I have my podcast on the left pane in Windows Terminal, dotnet monitor collect in the upper right, and the output lower right.

Image may be NSFW.
Clik here to view.

Once I figure out my process id (PID) - which will be automatic within a container as there will only be one - I can explore any of these local endpoints:

• /processes
• /dump/{pid?}
• /gcdump/{pid?}
• /trace/{pid?}
• /logs/{pid?}
• /metrics

If you are getting the logs, you'll get a never ending text/event-stream in your browser. I'd recommend you "curl" to see this at the command line.

Image may be NSFW.
Clik here to view.

You can also get momentary traces, collect a nettrace file and analyze it in Visual Studio, PerfView or other tools.

The dotnet monitor is experimental, but if you're digging it, head over to the dotnet/diagnostics GitHub and show some support.

SOS and Other Diagnostic Tools

• SOS - About the SOS debugger extension.
• dotnet-dump - Dump collection and analysis utility.
• dotnet-gcdump - Heap analysis tool that collects gcdumps of live .NET processes.
• dotnet-trace - Enable the collection of events for a running .NET Core Application to a local trace file.
• dotnet-counters - Monitor performance counters of a .NET Core application in real time.

Hope this helps you!

Sponsor: Upgrade from file systems and SQLite to Actian Zen Edge Data Management. Higher Performance, Scalable, Secure, Embeddable in most any programming language, OS, on 64-bit ARM/Intel Platform.