-
Notifications
You must be signed in to change notification settings - Fork 16.4k
Closed
Labels
area:providersgood first issuekind:bugThis is a clearly a bugThis is a clearly a bugprovider:amazonAWS/Amazon - related issuesAWS/Amazon - related issues
Description
Apache Airflow Provider(s)
amazon
Versions of Apache Airflow Providers
8.7.1
Apache Airflow version
2.7.2
Operating System
amazon linux AMI
Deployment
Amazon (AWS) MWAA
Deployment details
No response
What happened
The SageMakerTransformOperator deduplicates job name before creating the job, but does not deduplicate the model name before creating the model. Upon retry, the operator would fail due to model already exist error.
What you think should happen instead
The operator should apply the same behavior (fail or add timestamp) to both the job name and the model name.
How to reproduce
create a DAG containing a SageMakerTransformOperator step. In the operator config, fix both ModelName and TransformJobName. In airflow UI, run the task until completion in airflow once, and the run the same task for the second time by "clear task".
Anything else
No response
Are you willing to submit PR?
- Yes I am willing to submit a PR!
Code of Conduct
- I agree to follow this project's Code of Conduct
Metadata
Metadata
Assignees
Labels
area:providersgood first issuekind:bugThis is a clearly a bugThis is a clearly a bugprovider:amazonAWS/Amazon - related issuesAWS/Amazon - related issues