Azure Functions Linux EP1 deployment reports Oryx success but Python application content is not activated

Address in Estonia 0 Reputation points
2026-08-14T10:56:30.8266667+00:00

I need assistance with an Azure Functions platform deployment/content-activation issue that appears to require backend telemetry.

Environment:

Azure Functions Python programming model v2

Linux Elastic Premium EP1

Germany West Central

Remote Oryx build from a source-only ZIP package

SCM basic authentication disabled

Microsoft Entra authentication used for deployment evidence

Expected inventory: five functions defined in function_app.py

Observed behaviour:

The source package passed local integrity and structure validation.

The remote deployment command returned success.

Azure deployment records reported successful deployment.

Oryx reported a successful build.

The Function App was Running and the host appeared healthy.

No Python worker-startup exception was found.

Nevertheless, the active application content did not contain the expected current-generation function_app.py during a bounded 900-second readback window.

The authoritative function inventory remained empty instead of the expected five functions.

The controller stopped before restart, trigger synchronization, endpoint invocation, or any business operation. The temporary Function App, EP1 plan, and storage account were then removed. Sanitized deployment classifications, hashes, timestamps, and correlation information were retained and can be provided privately.

The production Function App selected in the Azure support wizard is unrelated and was selected only because the portal required an existing resource. Please do not modify or troubleshoot that production resource.

Questions:

How can a successful remote Oryx deployment fail to activate the resulting application content in the active site?

Is there a known Linux EP1 or Germany West Central deployment/site-specialization issue that could produce this state?

Can an Azure engineer review the backend deployment and content-activation telemetry?

  1. Because the required diagnostic identifiers should not be posted publicly, please escalate this to a private support case and provide a secure channel for the sanitized evidence.I need assistance with an Azure Functions platform deployment/content-activation issue that appears to require backend telemetry. Environment:
    • Azure Functions Python programming model v2
    • Linux Elastic Premium EP1
    • Germany West Central
    • Remote Oryx build from a source-only ZIP package
    • SCM basic authentication disabled
    • Microsoft Entra authentication used for deployment evidence
    • Expected inventory: five functions defined in function_app.py
    Observed behaviour:
    1. The source package passed local integrity and structure validation.
    2. The remote deployment command returned success.
    3. Azure deployment records reported successful deployment.
    4. Oryx reported a successful build.
    5. The Function App was Running and the host appeared healthy.
    6. No Python worker-startup exception was found.
    7. Nevertheless, the active application content did not contain the expected current-generation function_app.py during a bounded 900-second readback window.
    8. The authoritative function inventory remained empty instead of the expected five functions.
    The controller stopped before restart, trigger synchronization, endpoint invocation, or any business operation. The temporary Function App, EP1 plan, and storage account were then removed. Sanitized deployment classifications, hashes, timestamps, and correlation information were retained and can be provided privately. The production Function App selected in the Azure support wizard is unrelated and was selected only because the portal required an existing resource. Please do not modify or troubleshoot that production resource. Questions:
    1. How can a successful remote Oryx deployment fail to activate the resulting application content in the active site?
    2. Is there a known Linux EP1 or Germany West Central deployment/site-specialization issue that could produce this state?
    3. Can an Azure engineer review the backend deployment and content-activation telemetry?
    4. Because the required diagnostic identifiers should not be posted publicly, please escalate this to a private support case and provide a secure channel for the sanitized evidence.
Azure Functions
Azure Functions

An Azure service that provides an event-driven serverless compute platform.

0 comments

1 answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. Allan Solomon Mejia 3,510 Reputation points
    2026-08-14T19:04:12.4+00:00

    Hello @Address in Estonia

    Based on the details provided, I agree that this doesn't look like a typical Python/Oryx build failure. Your deployment completed successfully, Oryx reported success, and the Function host was running without a Python worker startup exception, but the deployed function_app.py wasn't present in the active application content and the function inventory remained empty.

    For Linux Elastic Premium, remote build is supported. Microsoft documents that remote build on Linux uses Oryx and requires:

    ENABLE_ORYX_BUILD=true

    SCM_DO_BUILD_DURING_DEPLOYMENT=true

    Microsoft also notes that Linux Premium apps have an SCM/Kudu site, although its capabilities are more limited than on Windows.

    One configuration I would verify is WEBSITE_RUN_FROM_PACKAGE. Microsoft specifically states that remote builds don't occur when an application is configured to use run-from-package in the conflicting deployment mode. Therefore, confirm that the deployment isn't mixing a remote Oryx build with an incompatible WEBSITE_RUN_FROM_PACKAGE configuration.

    That said, your evidence suggests the failure occurs after the build, during deployment/content activation or site specialization. A successful Oryx build only establishes that the source was successfully built; it doesn't by itself prove that the resulting artifact became the content being served by the active Function App.

    Given that:

    • the expected five functions never appeared,
    • the current function_app.py wasn't visible in the active content,
    • the deployment/build reported success,
    • no Python worker startup failure was observed, and
    • you already retained timestamps and correlation information,

    I think Azure Support/backend investigation is appropriate rather than repeatedly redeploying or modifying the unrelated production Function App.

    I would provide Support with the Function App resource ID, deployment timestamp, deployment/correlation IDs, Oryx deployment ID, region (Germany West Central), and the expected versus observed content hashes. Those should give the Functions engineering team enough information to correlate the deployment with platform-side content activation/site-specialization telemetry.

    I also wouldn't conclude yet that this is a known Germany West Central or EP1-specific issue without Microsoft confirming it from backend telemetry.

    Sharing you this reference:

    Microsoft Learn – Deployment technologies in Azure Functions

    https://learn.microsoft.com/en-us/azure/azure-functions/functions-deployment-technologies?

    Since the temporary Function App has already been removed, the retained correlation IDs and timestamps will be particularly important for any backend investigation.

    Please "Accept the Answer" if this information helped you. This will help us and others in the community as well.

    Was this answer helpful?

    0 comments

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.