Hello, I have a bug with kedro_azureml.dataset.AzureMLAssetDataset in the kedro-azureml plugin. It seems to be related to AzureMachineLearningFilesytem and this issue https://github.com/Azure/azure-sdk-for-python/issues/37089 Someone succeed to use azuremldataset in version 0.9.0 ?

can you share more details about the bug you are facing

here is my catalog :

azureml_companies:
    type: kedro_azureml.datasets.AzureMLAssetDataset
    azureml_dataset: kedro_companies_folder
    azureml_type: uri_folder
    root_dir: data/01_raw/test/
    dataset:
        type: pandas.CSVDataset
        filepath: "companies.csv"

Hi Adrien, just to clarify - are you trying to run your Kedro pipeline locally while your DataCatalog includes an

AzureMLAssetDataset

? Typically, the

kedro-azureml

plugin is used to convert and run Kedro pipelines within Azure ML itself. Have you tried executing it in an Azure ML compute environment, and does it work correctly there?

Hello, the error message is locally but i also tested it inside azureml

In this case, I’d recommend the following: 1. Debug locally using local copies of your data: Accessing

AzureMLAssetDataset

from a local machine can be tricky due to authentication issues. If you just want to debug your pipeline locally, consider downloading the datasets manually and temporarily updating your catalog to point to local file paths instead. 2. Check IAM permissions for Azure ML Compute: If you're seeing the same access error even when running inside Azure ML, it's likely that your compute instance doesn't have the correct permissions. Go to the Azure Portal and make sure the compute instance has a reader role under Access Control (IAM) for the relevant storage account.

Thanks for answer ! My pipeline is working locally with local csv dataset. I succed to use the dataset with a full azure pipeline. I'm evaluating if kedro could be a good add to our team. So it's not a permission problem ...

Yes, from what you've shared, the issue seems limited to accessing Azure Storage from your local environment - and that's seems unfortunately a common challenge with Azure’s SDK, rather than a Kedro-specific problem. Glad to hear it's working well in the Azure ML pipeline context!

Nono when i run with kedro in azureml it's crashing. What i mean was : i want to mirgate a pipeline from pure azureml to kedro with azureml plugin

what error do you see in the azure itself?

StreamAccess.NotFound Connection failed when trying to access the stream. Make sure uri is correct. error trying to connect: dns error: failled to lookup address information: Name or service not known

I see - it's hard to say for certain, but it looks like the issue is related to the dataset path or URI. I’d recommend double-checking the data asset path and how it’s referenced in your catalog.

mmmh i checked it 10 times ... Are you guys able to use this class on azureml or local ? If yes I would love to have an exemple !