Raise exception when provisioning fails

[JackCaster]

Currently, no error is thrown when the schedule fails to provision. Only a message is printed in the console 'Provisioning failed, please try again or contact support.':

# schedule.py
def _wait_for_provisioning(self, timeout_seconds):
        """
        Wait for the completion of provisioning for this schedule.

        Returns the status after the wait.

        :param timeout_seconds: Number of seconds to wait before timing out.
        :type timeout_seconds: int

        :return: The final status.
        :rtype: str
        """
        status = self._schedule_provider.get_schedule_provisioning_status(self.id)
        time_run = 0
        sleep_period = 5
        while status == 'Provisioning':
            if time_run + sleep_period > timeout_seconds:
                print('Timed out of waiting, status:%s.' % status, flush=True)
                break
            time_run += sleep_period
            time.sleep(sleep_period)
            status = self._schedule_provider.get_schedule_provisioning_status(self.id)
            print('Provisioning status:', status)
        if status == "Failed":
            print('Provisioning failed, please try again or contact support.')
        return status

If the provisioning fails, it is easy for the user to miss that. Also, there is no way the process can be safeguarded in production environments (e.g., stop the process if the schedule has not being properly provisioned). On top of this, when the provisioning fails, the schedule will still appear as Active (see also https://github.com/MicrosoftDocs/azure-docs/issues/77089).

---

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

• ID: bdf111d3-ff02-238f-68d9-5a36bb3bf75b
• Version Independent ID: da555131-26d9-c768-d8d4-e46c485d2601
• Content: azureml.pipeline.core.schedule.Schedule class - Azure Machine Learning Python
• Content Source: AzureML-Docset/stable/docs-ref-autogen/azureml-pipeline-core/azureml.pipeline.core.schedule.Schedule.yml
• Service: machine-learning
• Sub-service: core
• GitHub Login: @DebFro
• Microsoft Alias: debfro

[shbijlan]

@JackCaster thank you for the feedback. I have reported this issue to engineering and will update after investigation here.