App Service Kudu Profiler fails with 429 “concurrent profiling sessions” even after restart/swap

Vibe FYI 0 Reputation points
2026-10-04T02:12:30.6933333+00:00

We are seeing a persistent profiling failure on our Azure App Service (Windows, ASP.NET Core on IIS).

Error Starting a profiling session now would exceed the limit of concurrent profiling sessions. Please try again in a few minutes. Also seen: Failed to profile process <pid>. Profiling call failed with 429.

What is happening

  1. Kudu Process Explorer profiling cannot be started.
  2. “Collect a Profiler Trace” in portal also fails with the same 429.
  3. This used to work previously on the same app.

What we already tried

  1. Application restart and advanced application restart.
  2. Deployment and slot swap.
  3. Retrying from Kudu Process Explorer.
  4. Restarting DaaS WebJob.
  5. Waiting and retrying later.

Result Still fails with the same 429 concurrent profiling sessions message.

Question

  1. Is this usually a stale DaaS/profiler lease at App Service Plan level?
  2. Any additional non-destructive checks we can do from customer side?
  3. If this is platform-side, what is the recommended support path to clear the stuck profiling lock?

Additional context

  1. Single app instance currently visible.
  2. No active profiling session visible in Kudu UI when trying again.
Azure App Service
Azure App Service

Azure App Service is a service used to create and deploy scalable, mission-critical web apps.

0 comments

1 answer

Sort by: Most helpful
Most helpful Newest Oldest

  1. AI answer

    2026-10-04T02:12:48.86+00:00

    Yes. This error is consistent with App Service plan/VM-level contention, not only an app-level session visible in Kudu.

    Azure App Service allows only one profiling session at a time per VM. That limit is enforced across all apps and deployment slots in the same App Service plan. It applies to sessions started from Diagnose and solve problems, Kudu, and Application Insights Profiler for .NET. If another session is already running on the same VM, the profiler can fail to start and log errors such as:

    • Microsoft.ServiceProfiler.Exceptions.TooManyETWSessionException
    • StartProfiler failed ... HRESULT: 0xE111005E

    A visible absence of an active session in the Kudu UI for this app does not rule out contention from another app or slot sharing the same plan/VM.

    Additional non-destructive checks from the customer side:

    1. Check whether other apps or deployment slots share the same App Service plan.
      • Running slots also participate in profiler contention.
      • If unused slots exist, stop them.
    2. Disable Profiler on apps that do not need profiling.
      • This reduces contention and log noise.
    3. If profiling is enabled on staging slots, consider temporarily disabling it during deployments or slot swaps.
      • Slot activity can trigger concurrent sessions.
    4. Open the Diagnostic Services site extension status page.
      • In Azure portal, open the App Service.
      • Select Advanced Tools > Go.
      • In the Kudu site, append /DiagnosticServices to the URL.
      • Review installation state and any unexpected error shown for the .NET Profiler or Snapshot Debugger agents.
    5. If needed, stop the ApplicationInsightsProfiler3 WebJob for the specific app instance.
      • In Azure portal, go to the App Service.
      • Under Settings, select WebJobs.
      • Select ApplicationInsightsProfiler3.
      • Select Stop and confirm.
    6. If the app is on a shared or free App Service plan, Profiler is not supported there. Use at least a Basic plan.

    If the issue persists after checking shared apps/slots and the Diagnostic Services status page, the supported mitigation in the documentation is to reduce VM sharing pressure:

    • Move some web apps to a different App Service plan.
    • Stop unused deployment slots.
    • Disable Profiler where it is not needed.

    The provided documentation does not describe a customer-side action to clear a platform-side stuck profiling lock beyond those checks and mitigations.

    References:

    AI-generated content may be incorrect. Read our transparency notes for more information.

    Was this answer helpful?

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.