How to drop front door health probes from application insights using agent on app service

Matt Browne 21 Reputation points
2021-07-02T08:05:55.527+00:00

We are using azure app service codeless implementation of application insights: https://learn.microsoft.com/en-us/azure/azure-monitor/app/azure-web-apps?tabs=net#enable-agent-based-monitoring

We are also using front door, therefore all the health prob HEAD requests are ending up in application insights creating a lot of noise and extra cost.

I understand if you are using the application insights SDK and have an applicationinsights.config file you can filter these requests out.

But is there a way of doing this using the agent based monitoring, the doc hints that applicationinsights.config settings can be set as application settings in the app service, but does anyone have an example of how to do filtering this way?

Azure Monitor
Azure Monitor
An Azure service that is used to collect, analyze, and act on telemetry data from Azure and on-premises environments.
2,783 questions
Azure Front Door
Azure Front Door
An Azure service that provides a cloud content delivery network with threat protection.
573 questions
0 comments

1 answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. BhargaviAnnadevara-MSFT 5,446 Reputation points
    2021-08-09T13:32:00.587+00:00

    @Matt Browne Thanks for reaching out. Currently, Telemetry processors (preview) feature for filtering out unwanted telemetry data is available only for codeless application monitoring for Java apps via the standalone Java 3.x agent (examples here).

    For other environments/languages and advanced configurations, manual instrumentation with the SDK might still be the way to go. Although it would require some management effort, this approach is much more customizable and would give you greater control over the telemetry you want to ingest.

    Regardless of the approach, to reduce the volume of telemetry without affecting your statistics, you can try configuring Sampling, either via Application settings or the SDK.

    From a Front Door configuration perspective, you could increase the Interval between health checks to reduce the frequency of requests. Or, if you have a single backend in your backend pool, you can choose to disable the health probes. Even if you have multiple backends in the backend pool but only one of them is in enabled state, you can disable health probes. This should help in reducing the load on your application backend, and also the telemetry traffic.

    Hope this helps. Do let us know if you have further questions.

    ----------

    If an answer is helpful, please "Accept answer" and/or "Up-Vote" which might help other community members reading this thread.

    0 comments