Search code examples
azure-machine-learning-service

Workaround for timeout error in Dataset.download()


azureml-sdk version: 1.0.85

Calling below (as given in the Dataset UI), I get this

ds_split = Dataset.get_by_name(workspace, name='ret- holdout-split')
ds_split.download(target_path=dir_outputs, overwrite=True)
UnexpectedError:
{'errorCode': 'Microsoft.DataPrep.ErrorCodes.Unknown', 'message':
    'The client could not finish the operation within specified timeout.',
    'errorData': {}}

The FileDataset 1GB pickled file stored in blob. Here's a gist with the full traceback


Solution

  • Tried again this AM and it worked. let's file this under "transient error"