Skip to content

Diagnostics

Diagnostics provides a set of features that make it easy for us to document critical operations that occurs during the application's operation, their execution time, etc., allowing administrators to find the root cause of problems, especially in production environments.

Tracing

The CAP provides support for DiagnosticSource with a listener name of CapDiagnosticListener.

Diagnostics provides tracing event information as follows:

  • Before the message is persisted
  • After the message is persisted
  • Message persistence exception
  • Before the message is sent to MQ
  • After the message is sent to MQ
  • The message sends an exception to MQ.
  • Messages saved from MQ consumption before saving
  • After the message is saved from MQ consumption
  • Before the subscriber method is executed
  • After the subscriber method is executed
  • Subscriber method execution exception

Related objects, you can find at the DotNetCore.CAP.Diagnostics namespace.

Tracing with Apache Skywalking

Skywalking's C# client provides support for CAP Diagnostics. You can use SkyAPM-dotnet to tracking.

Try to read the README to integrate it in your project.

Example tracking image :

Others APM support

There is currently no support for APMs other than Skywalking, and if you would like to support CAP diagnostic events in other APM, you can refer to the code here to implement it:

At present, apart from Skywalking, we have not provided support for other APMs. If you need it, you can refer the code here to implementation, and we also welcome the Pull Request.

https://github.com/SkyAPM/SkyAPM-dotnet/tree/master/src/SkyApm.Diagnostics.CAP

Metrics

Metrics are numerical measurements reported over time, most often used to monitor the health of an application and generate alerts. For example, a web service might track how many requests it receives each second, how many milliseconds it took to respond, and how many of the responses sent an error back to the user.

CAP 7.0 is support for EventSource, and the counters name is DotNetCore.CAP.EventCounter.

CAP provides the following metrics:

  • Publish rate pre seconds
  • Consume rate pre seconds
  • Invoke Subscriber rate pre seconds
  • Subscriber elpased time mean pre seconds

Monitor with dotnet-counters

dotnet-counters is a performance monitoring tool for ad-hoc health monitoring and first-level performance investigation. It can observe performance counter values that are published via the EventCounter API or the Meter API.

Use the following commands to monitor metrics in CAP:

dotnet-counters ps
dotnet-counters monitor --process-id=25496 --counters=DotNetCore.CAP.EventCounter

process-id： The ID of the CAP process to collect counter data from.

img

Monitor with dashboard

You can configure x.UserDashboard() to open the dashboard to view Metrics graph charts.

img

In the Realtime Metric Graph, the time axis will scroll in real time over time so that you can see the rate of publishing and consuming messages per second, And the consumer execution time is "dotted" on the Y1 axis (Y0 axis is the rates, and the Y1 axis is the execution elpsed time).