問題文
Your team must be able to recreate an identical Azure Machine Learning workspace, including its datastores and compute clusters, in a new subscription for disaster recovery drills. The definition must be reviewed and versioned in the same Git repository as the training code. Which approach best meets this requirement?
選択肢
- Take a backup of the workspace storage account and restore it into the new subscription
- Author Bicep templates for the workspace and its resources, and apply them with the Azure CLI from a versioned Git repository
- Export the workspace assets to a notebook and rerun the notebook cells manually
- Create the workspace in the portal and document each screen in a runbook