問題文
You need to create an Azure Machine Learning compute cluster that costs nothing when no jobs are running, but can grow to eight nodes when several jobs are submitted. Which configuration achieves this?
選択肢
- Attach a Kubernetes cluster with eight nodes and rely on its own autoscaler, because the workspace is charged only while a job occupies one of those nodes
- Set the minimum number of nodes to zero and the maximum number of nodes to eight
- Create eight separate compute instances and start them manually when needed
- Set the minimum number of nodes to eight and enable idle shutdown