Endless loop when try to deploy a sagemaker pipeline from a windows machine

[StefanHemetsberger]

Describe the bug When trying to deploy a sagemaker pipeline from the local machine (windows) to the cloud, an infinite loop will happen.

To reproduce Must be a Windows Machine.

Imports: from sagemaker.estimator import Estimator from sagemaker.workflow.pipeline import Pipeline

Code: estimator = Estimator(...) train_args = estimator.fit(...) step_train = TrainingStep(...) pipeline = Pipeline(...) pipeline.upsert(...) --> will stop here an wait endless

The while loop here: src/sagemaker/_studio.py in _find_config() waits because there is no STUDIO_PROJECT_CONFIG available.

Expected behavior Upload pipeline works and be able to start pipeline execution.

Screenshots or logs A ... never be true because root path is "E:/" B ... working directory will never change to "/" to be able to escape the while loop

System information A description of your system. Please provide:

• SageMaker Python SDK version: 2.226.1
• Framework name (eg. PyTorch) or algorithm (eg. KMeans): -
• Framework version: -
• Python version: 3.12.0
• CPU or GPU: CPU
• Custom Docker image (Y/N): Y

Additional context If manually skip the while-loop in "src/sagemaker/_studio.py" _find_config(), the pipeline will start executing as expected

[qidewenwhen]

Hi @StefanHemetsberger , thanks for using SageMaker Pipeline!

I'm from the SageMaker Pipeline team.

Unfortunately, the entire SageMaker Python SDK does not support Windows OS. Please find the Supported Operating System information in https://github.com/aws/sagemaker-python-sdk/tree/master?tab=readme-ov-file#supported-operating-systems

So currently this Python SDK has to be run in a Unix/Linux/Mac OS.

I'll change the type of this issue from bug to feature request and let SageMaker Python SDK to keep track of it.

[StefanHemetsberger]

Alright, thanks for your answer.