Azure Machine Learning Compute Instance Data "Mount" option unavailable in studio

Question

In the compute instance management interface there is a 'Data (preview)' tab that I've previously used to mount a datastore to a compute instance. In newly-created compute instances the Mount option is unavailable (see below). Are there requirements on the compute instance settings for this option to be available? I haven't been able to find any documentation on this preview feature. Is there a CLI method for creating these mounts?

Answer 1

Hi Sam Preston,

Thanks for reaching out to Microsoft Q&A.

The below code snippet demonstrates how to mount a datastore in Azure Machine Learning using Python. The mount method creates a context that must be started to access the mounted files.

from azureml.core import Workspace, Datastore, Dataset
# Connect to the workspace
workspace = Workspace.from_config()
# Retrieve the datastore
datastore = Datastore.get(workspace, 'your_datastore_name')
# Create a dataset from the datastore
dataset = Dataset.File.from_files((datastore, 'path/to/your/files'))
# Mount the dataset
mounted_path = '/mnt/mount_dir'
mount_context = dataset.mount(mounted_path)
# Start the mount context
mount_context.start()
# Use the mounted path for your operations
# Stop the mount context when done
mount_context.stop()

Please 'Upvote'(Thumbs-up) and 'Accept' as an answer if the reply was helpful. This will benefit other community members who face the same issue.