Azure Function App (Custom Docker Container) fails to start consistently – recurring container startup timeout / portal visibility issue

Question

We are experiencing a recurring startup failure with Azure Function Apps deployed using custom Docker images from Docker Hub via the Deployment Center (Container Image).

This is not a new issue – it is the same behavior we previously reported to Microsoft Support in an earlier ticket, which was temporarily mitigated by changing platform settings.

Environment

1. Platform: Azure Function app (Linux)
2. Deployment type: Custom container image (Docker Hub)
3. Hosting: Azure App Elastic premium (EP1) Service Plan
4. Deployment method: Deployment Center → Docker image

Issue Summary:

When creating a new Function App with a custom container image, the Function App:

1. Fails to start on initial deployment
2. Does not respond to health checks within the default timeout
3. Does not appear properly in the Azure Portal UI in some cases
4. Does not respond to HTTP triggers unless manually mitigated

This behavior is consistent and reproducible across new Function Apps.

Previous Microsoft Support Case (Microsoft Q&A thread discussion - 5566336)

We previously reported the same issue to Microsoft Support.

In that case, Microsoft Support identified the following:

1. The container started but failed health checks within the default 230 seconds
2. The issue was mitigated by manually adding the application setting: WEBSITES_CONTAINER_START_TIME_LIMIT = 1800
3. After adding this setting, the Function App came online and began returning 200 OK responses.

Microsoft Support at the time stated this could be a portal-side (UI) platform issue that required investigation by the product team, and mentioned that an RCA (Root Cause Analysis) would be provided.

However, we never received the RCA, and the issue has now reappeared in a new Function App, confirming this is persistent and platform-related, not an isolated incident.

Answer 1

It seems you are encountering a recurring startup failure with your Azure Function App that uses a custom Docker image. Based on the information provided, the following points are relevant:

1. Startup Timeout: The default startup timeout for Azure Function Apps is 230 seconds. If your container fails to respond to health checks within this time frame, it will result in a startup failure. You mentioned that adding the application setting WEBSITES_CONTAINER_START_TIME_LIMIT to 1800 seconds previously mitigated the issue, which indicates that your container may need more time to start up due to its configuration or resource requirements.
2. Health Checks: Ensure that your container is configured to respond to health checks correctly. If the application inside the container is not ready to handle requests within the specified timeout, it will lead to the issues you're experiencing.
3. Portal Visibility Issues: The problems with the Azure Portal UI not displaying the Function App correctly could be related to the startup failures. If the app is not starting properly, it may not register correctly in the portal.
4. Manual Mitigation: Since the issue is consistent across new Function Apps, and you've previously mitigated it by adjusting the startup time limit, it may be worth applying this setting again as a temporary fix while awaiting further investigation from Microsoft Support.
5. Root Cause Analysis: It appears that there has been a lack of follow-up from Microsoft Support regarding the root cause analysis (RCA) for this issue. It may be beneficial to re-engage with support to seek updates on the RCA and any potential fixes or recommendations they may have.

In summary, adjusting the WEBSITES_CONTAINER_START_TIME_LIMIT setting seems to be a necessary step to allow your Function App to start successfully. Additionally, ensuring that your container is properly configured to respond to health checks is crucial for resolving these startup issues.

---

References:

• Troubleshoot error: "Azure Functions Runtime is unreachable"
• Troubleshoot start failures in Azure Container Apps

Answer 2

I’ve actually seen this exact behavior with custom Docker-based Function Apps on Elastic Premium. The container itself starts, but it never clears the health check window, so the platform keeps recycling it.

In our case the Function App also disappeared from the portal or showed a blank overview, just like you described. It definitely felt like a platform-side issue rather than anything wrong with the image.

The only reliable workaround we found was the same one support gave you earlier.

Extending WEBSITES_CONTAINER_START_TIME_LIMIT to something much higher (we used 1800 seconds) let the app fully boot and finally respond. Once that setting was in place, everything behaved normally.

If the issue came back on new Function Apps even after the previous ticket, then you’re not imagining it. This is one of those cases where the platform struggles with slower-booting custom images on EP plans. I’d still push Support for an RCA because this isn’t expected behavior.

In short, nothing on your side looks misconfigured. It’s the platform retrying too aggressively, and increasing the startup time limit is the only consistent fix we’ve used until Microsoft addresses it properly.