An example for a customer on how to use Splunk APM on a .NET application running in Kubernetes.
The Dockerfile in this repo is used to load the signalfx-dotnet-tracing library files to their default directory (/opt/signalfx-dotnet-tracing).
Then, it's used as an initContainer in the example.yaml manifest so that the files are then available via a volumeMount in the application container.
Then all that's left is to set the appropriate environment variables for the tracer (see: https://github.com/signalfx/signalfx-dotnet-tracing#configure-the-signalfx-tracing-library-for-net).
Note, I've enabled debugging on the tracer for this example so that you can see if it's working or not. You should see entries like this in your application container's logs:
[DBG] Span started: [s_id: 14938313515406949587, p_id: null, t_id: 1b3b0564434bc315b72ecc905a51a008]
[DBG] Span closed: [s_id: 14938313515406949587, p_id: null, t_id: 1b3b0564434bc315b72ecc905a51a008] for (Service: TestASPNetCoreService, Resource: /, Operation: /, Tags: [environment,signalfx.tracing.library,signalfx.tracing.version,component,span.kind,http.method,http.request.headers.host,http.url,http.status_code])
Also note that in my setup (how I've tested this), the spans are being sent to an OTEL collector running in my Kubernetes cluster.
To test this out, run:
$ kubectl apply -f example.yaml
Wait for a few seconds to make sure the application container is running. The application can then be reached by:
$ kubectl port-forward service/dotnettestservice 8080:8080
Then open up http://localhost:8080 in your browser to send some traffic to the app.