Use managed compute in a managed virtual network

Learn how to configure compute clusters or compute instances in an Azure Machine Learning managed virtual network.

When you use a managed network, compute resources managed by Azure Machine Learning can participate in the virtual network. Azure Machine Learning compute clusters, compute instances, and managed online endpoints are created in the managed network.

This article focuses on configuring compute clusters and compute instances in a managed network. For information on managed online endpoints, see secure online endpoints with network isolation.

Important

If you plan on using serverless Spark jobs, see the managed virtual network article for configuration information. These steps must be followed when configuring the managed virtual network.

Prerequisites

Before following the steps in this article, make sure you have the following prerequisites:

Azure CLI

• An Azure Machine Learning workspace configured to use a managed virtual network.

• The Azure CLI and the ml extension to the Azure CLI. For more information, see Install, set up, and use the CLI (v2).

  Tip

  Azure Machine Learning managed virtual network was introduced on May 23rd, 2023. If you have an older version of the ml extension, you may need to update it for the examples in this article work. To update the extension, use the following Azure CLI command:

  az extension update -n ml
  

• The CLI examples in this article assume that you're using the Bash (or compatible) shell. For example, from a Linux system or Windows Subsystem for Linux.

• The Azure CLI examples in this article use ws to represent the name of the workspace, and rg to represent the name of the resource group. Change these values as needed when using the commands with your Azure subscription.

Python SDK

• An Azure Machine Learning workspace configured to use a managed virtual network.

• The Azure Machine Learning Python SDK v2. For more information on the SDK, see Install the Python SDK v2 for Azure Machine Learning.

  Tip

  Azure Machine learning managed virtual network was introduced on May 23rd, 2023. If you have an older version of the SDK installed, you may need to update it for the examples in this article to work. To update the SDK, use the following command:

  pip install --upgrade azure-ai-ml azure-identity
  

• The examples in this article assume that your code begins with the following Python. This code imports the classes required when creating a workspace with managed virtual network, sets variables for your Azure subscription and resource group, and creates the ml_client:

  from azure.ai.ml import MLClient
  from azure.ai.ml.entities import (
      Workspace,
      ManagedNetwork,
      IsolationMode,
      ServiceTagDestination,
      PrivateEndpointDestination,
      FqdnDestination
  )
  from azure.identity import DefaultAzureCredential
  
  # Replace with the values for your Azure subscription and resource group.
  subscription_id = "<SUBSCRIPTION_ID>"
  resource_group = "<RESOURCE_GROUP>"
  workspace_name = "<WORKSPACE_NAME>"
  
  # get a handle to the subscription
  ml_client = MLClient(
      workspace_name=workspace_name,
      subscription_id=subscription_id,
      resource_group_name=resource_group,
      credential=DefaultAzureCredential()
  )
  

Studio

• An Azure Machine Learning workspace configured to use a managed virtual network.

Configure compute resources

Use the following tabs to learn how to configure compute clusters and compute instances in a managed virtual network:

Tip

When using a managed virtual network, compute clusters and compute instances are automatically created in the managed network. The following steps focus on configuring the compute resources to not use a public IP address.

Azure CLI

To create a compute cluster with no public IP, use the following command:

az ml compute create --name cpu-cluster --resource-group rg --workspace-name ws --type AmlCompute --set enable_node_public_ip=False

To create a compute instance with no public IP, use the following command:

az ml compute create --name myci --resource-group rg --workspace-name ws --type ComputeInstance --set enable_node_public_ip=False

Python SDK

The following Python SDK example shows how to create a compute cluster and compute instance with no public IP:

from azure.ai.ml.entities import AmlCompute

# Create a compute cluster
compute_cluster = AmlCompute(
    name="mycomputecluster",
    size="STANDARD_D2_V2",
    min_instances=0,
    max_instances=4,
    enable_node_public_ip=False
)
ml_client.begin_create_or_update(entity=compute_cluster)

# Create a compute instance
from azure.ai.ml.entities import ComputeInstance

compute_instance = ComputeInstance(
    name="mycomputeinstance",
    size="STANDARD_DS3_V2",
    enable_node_public_ip=False
)
ml_client.begin_create_or_update(compute_instance)

Studio

You can't create a compute cluster or compute instance from the Azure portal. Instead, use the following steps to create these computes from Azure Machine Learning studio:

1. From studio, select your workspace.

2. Select the Compute page from the left navigation bar.

3. Select the + New from the navigation bar of compute instance or compute cluster.

4. Configure the VM size and configuration you need, then select Next until you arrive at the following pages:

   • For a compute cluster, use the Advanced Settings page and select the No Public IP option to remove the public IP address.

     

   • For a compute instance, use the Security page and select the No Public IP option to remove the public IP address.

     

5. Continue with the creation of the compute resource.

Limitations

• Creating a compute cluster in a different region than the workspace isn't supported when using a managed virtual network.
• If the compute is in a managed network and also configured for no public IP, use the az ml compute connect-ssh command to connect to the compute instance over SSH.

Migration of compute resources

If you have an existing workspace and want to enable managed virtual network for it, there's currently no supported migration path for existing manged compute resources. You'll need to delete all existing managed compute resources and recreate them after enabling the managed virtual network. The following list contains the compute resources that must be deleted and recreated:

• Compute cluster
• Compute instance
• Managed online endpoints

Next steps

• Managed virtual network isolation
• Secure online endpoints with network isolation