How can I capture and report the specific reason for SageMaker pipeline failures?

Question

I have setup event bridge rules to receive SageMaker pipeline status notifications. However, I would like to inquire about how to obtain more detailed information about pipeline failures beyond just knowing that the pipeline has failed. Is there a way to capture and report the specific reason for a pipeline failure? I am looking forward to your guidance on how to implement this.
Thank you.

Answer

https://docs.aws.amazon.com/sagemaker/latest/dg/pipelines-troubleshooting.html
Please refer above link which summarized possible pipeline issues. Most of them will be fixed during dev time (for example, lack of permissions)

For execution script logs, they will be caputured in cloudwatch. 
https://docs.aws.amazon.com/sagemaker/latest/dg/pipelines-studio-view-execution.html

If you are more interested in build your own dashboard to view execution result, consider to use boto3 api to retrieve execution result,
https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/sagemaker/client/describe_pipeline_execution.html

How can I capture and report the specific reason for SageMaker pipeline failures?

相關內容