Deployment of prompt flow endpoint crashes across multiple Azure environments

Question

Andrew Burgess 25

Cannot currently deploy any prompt flow endpoints in Azure ML Studio. Error is:

ResourceOperationFailure: ResourceNotReady: User container has crashed or terminated

This occurs in different Azure environments so seems to be a general issue.

YutongTie-MSFT 53,966 Reputation points Moderator

2024-03-08T22:09:05.12+00:00

@Andrew Burgess Thanks for reaching out to us, seems like you were attempting to use custom container, but the environment definition was causing the service to believe this was not a custom container.

The working scenario was using CLI and yaml files from here: Deploy a registered R model to an online (real time) endpoint - Azure Machine Learning | Microsoft Learn

Need to update python SDKv2 code: Deploy a model in a custom container to an online endpoint - Azure Machine Learning | Microsoft Learn

Is that your case? If not, could you please share more details about your case? Thanks.

Regards,

Yutong
Andrew Burgess 25 Reputation points

2024-03-11T09:06:45.2433333+00:00

Hi Yutong, I'm not sure of any of the technical details as I just used the automatic job that runs when you click on the Deploy button from the Prompt Flow. It seems that there is an issue in the job that runs.
YutongTie-MSFT 53,966 Reputation points Moderator

2024-03-31T04:24:24.1633333+00:00

@Andrew Burgess No problem for having no technical details, I can enable you a free support ticket for your issue if you have no support plan. Please let me know if you need it.

Regards,

Yutong

1 answer

YutongTie-MSFT 53,966 Reputation points Moderator

2024-03-08T22:09:05.12+00:00

@Andrew Burgess Thanks for reaching out to us, seems like you were attempting to use custom container, but the environment definition was causing the service to believe this was not a custom container.

The working scenario was using CLI and yaml files from here: Deploy a registered R model to an online (real time) endpoint - Azure Machine Learning | Microsoft Learn

Need to update python SDKv2 code: Deploy a model in a custom container to an online endpoint - Azure Machine Learning | Microsoft Learn

Is that your case? If not, could you please share more details about your case? Thanks.

Regards,

Yutong
Andrew Burgess 25 Reputation points

2024-03-11T09:06:45.2433333+00:00

Hi Yutong, I'm not sure of any of the technical details as I just used the automatic job that runs when you click on the Deploy button from the Prompt Flow. It seems that there is an issue in the job that runs.
YutongTie-MSFT 53,966 Reputation points Moderator

2024-03-31T04:24:24.1633333+00:00

@Andrew Burgess No problem for having no technical details, I can enable you a free support ticket for your issue if you have no support plan. Please let me know if you need it.

Regards,

Yutong

Answer 1

Andrew Burgess 25

Since the time of my last post, the problem now seems to have been fixed. Must have been som error caused and rectified by Microsoft.