What causes Fabric Pipeline Error 4125?

Transient server-side failure or outage within the Azure Machine Learning service. Azure ML service is temporarily unavailable or under maintenance. Intermittent HTTP errors (e.g., 500, 503) returned by the Azure ML endpoint. High load or throttling on the Azure Machine Learning backend

How do I fix Fabric Pipeline Error 4125?

Step 1: Note the HTTP status code and external message from the error to determine the nature of the failure (e.g., 500 Internal Server Error vs. 503 Service Unavailable).. Step 2: Wait a few minutes and retry the pipeline run, as the error is likely transient.. Step 3: Check the Azure Service Health dashboard for any active incidents or degradations affecting Azure Machine Learning in your region.. Step 4: If the issue persists after multiple retries, open a support ticket with the Azure Machine Learning team, providing the operation name, status code, and external message from the error.. Step 5: Consider adding retry policies to your Fabric pipeline activity to automatically handle transient Azure ML errors.

High severityconnectivityMicrosoft Fabric →

Power BI Refresh Error:
Fabric Pipeline Error 4125

What does this error mean?

A request sent to Azure Machine Learning failed due to a server-side error on the Azure ML side. This is typically a transient infrastructure issue within the Azure Machine Learning service.

Common causes

1Transient server-side failure or outage within the Azure Machine Learning service
2Azure ML service is temporarily unavailable or under maintenance
3Intermittent HTTP errors (e.g., 500, 503) returned by the Azure ML endpoint
4High load or throttling on the Azure Machine Learning backend

How to fix it

1Step 1: Note the HTTP status code and external message from the error to determine the nature of the failure (e.g., 500 Internal Server Error vs. 503 Service Unavailable).
2Step 2: Wait a few minutes and retry the pipeline run, as the error is likely transient.
3Step 3: Check the Azure Service Health dashboard for any active incidents or degradations affecting Azure Machine Learning in your region.
4Step 4: If the issue persists after multiple retries, open a support ticket with the Azure Machine Learning team, providing the operation name, status code, and external message from the error.
5Step 5: Consider adding retry policies to your Fabric pipeline activity to automatically handle transient Azure ML errors.

Frequently asked questions

Should I always retry when I see Error 4125?

Yes, a single retry is warranted since this is a server-side Azure ML error. However, if the error persists beyond 2–3 retries, you should check Azure Service Health and contact Azure ML support rather than continuing to retry indefinitely.

Can I configure automatic retries for this error in Fabric pipelines?

Yes, Fabric pipeline activities support retry configuration. Set a retry count of 2–3 with an appropriate retry interval (e.g., 60–120 seconds) on the Azure ML activity to handle transient server errors automatically.