Closed mvbugge closed 2 years ago
That's correct. When invoking an AML pipeline, the variable SERVICECONNECTION_WS
is used instead of SERVICECONNECTION_RG
(both are defined in the infra configuration file). For example in the modeling pipeline: https://github.com/microsoft/dstoolkit-mlops-base/blob/322f45171b91646eae8b4d276c371d3685d7c54f/azure-pipelines/PIPELINE-1-modeling.yml#L97
If the wrong service connection is used, the task will fail with no explanation. That issue is documented in the FAQ here.
I'll close the issue now, thanks for creating it so people will find the solution in the future.
In the pipeline azure-pipelines/templates/utils/invoke-aml-pipeline.template.yml
`
Reading online, and from my own experiences, the task will not run unless given a specific machine learning workspace service principal connection in DevOps. A typical service principal connection will not suffice.
No error message is available, hence this issue needs to be documented