Executing pipeline in AML from ADF suddenly stopped working

Question

I have a pipeline defined in Azure Machine Learning. It was launched every day with Azure Data Factory with Machine Learning Execute Pipeline activity. This solution worked without any issues for a few weeks, but since 12/09/2021 all pipeline runs have failed with error: User starting the run is not an owner or assigned user to the Compute Instance.
I did not change anything in ADF or AML.

Should I assign compute to ADF? How to do this?

Answer 1

I ran into this same issue in a slightly different context. I didn't manage to figure out the root cause but managed to resolve it in practice by standing up a Compute Cluster instead of a Compute Instance (see https://learn.microsoft.com/en-us/azure/machine-learning/how-to-create-attach-compute-cluster?tabs=python)

Answer 2

Using compute cluster as compute target fix the issue.
Root cause is compute instance is assigned to certain user or team, and service principle or managed identity is not in the assigned user group in compute instance. Compute cluster can be used by multi-users, and other services.