RemoteCompute Class

Manages a remote compute target for use in Azure Machine Learning.

Azure Machine Learning supports using attaching a remote compute resource to your workspace. The remote resource can can be an Azure VM, a remote server in your organization, or on-premises, as long as the resource is accessible to Azure Machine Learning. For more information, see What are compute targets in Azure Machine Learning?

Class ComputeTarget constructor.

Retrieve a cloud representation of a Compute object associated with the provided workspace. Returns an instance of a child class corresponding to the specific type of the retrieved Compute object.

Inheritance
RemoteCompute

Constructor

RemoteCompute(workspace, name)

Parameters

Name Description
workspace
Required

The workspace object containing the RemoteCompute object to retrieve.

name
Required
str

The name of the of the RemoteCompute object to retrieve.

workspace
Required

The workspace object containing the Compute object to retrieve.

name
Required
str

The name of the of the Compute object to retrieve.

Remarks

The following Azure regions do not support using the public IP address of a virtual machine or HDInsight cluster to attach the compute target.

  • US East

  • US West 2

  • US South Central

Instead, use the Azure Resource Manager ID of the VM or HDInsight cluster. The resource ID of the VM can be constructed using the subscription ID, resource group name, and VM name using the following string format: /subscriptions/<subscription_id>/resourceGroups/<resource_group>/providers/Microsoft.Compute/virtualMachines/<vm_name>.

The following example show how to create and attach a Data Science Virtual Machine (DSVM) as a compute target.


   from azureml.core.compute import ComputeTarget, RemoteCompute
   from azureml.core.compute_target import ComputeTargetException

   username = os.getenv('AZUREML_DSVM_USERNAME', default='<my_username>')
   address = os.getenv('AZUREML_DSVM_ADDRESS', default='<ip_address_or_fqdn>')

   compute_target_name = 'cpudsvm'
   # if you want to connect using SSH key instead of username/password you can provide parameters private_key_file and private_key_passphrase
   try:
       attached_dsvm_compute = RemoteCompute(workspace=ws, name=compute_target_name)
       print('found existing:', attached_dsvm_compute.name)
   except ComputeTargetException:
       attach_config = RemoteCompute.attach_configuration(address=address,
                                                          ssh_port=22,
                                                          username=username,
                                                          private_key_file='./.ssh/id_rsa')


   # Attaching a virtual machine using the public IP address of the VM is no longer supported.
   # Instead, use resourceId of the VM.
   # The resourceId of the VM can be constructed using the following string format:
   # /subscriptions/<subscription_id>/resourceGroups/<resource_group>/providers/Microsoft.Compute/virtualMachines/<vm_name>.
   # You can also use subscription_id, resource_group and vm_name without constructing resourceId.
       attach_config = RemoteCompute.attach_configuration(resource_id='<resource_id>',
                                                          ssh_port=22,
                                                          username='username',
                                                          private_key_file='./.ssh/id_rsa')

       attached_dsvm_compute = ComputeTarget.attach(ws, compute_target_name, attach_config)

       attached_dsvm_compute.wait_for_completion(show_output=True)

Full sample is available from https://github.com/Azure/MachineLearningNotebooks/blob/master/how-to-use-azureml/training/train-on-remote-vm/train-on-remote-vm.ipynb

Methods

attach

DEPRECATED. Use the attach_configuration method instead.

Associate an existing remote compute resource with the provided workspace.

attach_configuration

Create a configuration object for attaching a remote compute target.

Attaching a virtual machine using the public IP address of the VM is no longer supported. Instead, use the resourceId of the VM. The resourceId of the VM can be constructed using the following string format: "/subscriptions/<subscription_id>/resourceGroups/<resource_group>/ providers/Microsoft.Compute/virtualMachines/<vm_name>".

You can also use subscription_id, resource_group and vm_name without constructing resourceId. For more information, see https://aka.ms/azureml-compute-vm.

delete

Delete is not supported for a RemoteCompute object. Use detach instead.

deserialize

Convert a JSON object into a RemoteCompute object.

detach

Detach the RemoteCompute object from its associated workspace.

Underlying cloud objects are not deleted, only the association is removed.

get_credentials

Retrieve the credentials for the RemoteCompute target.

refresh_state

Perform an in-place update of the properties of the object.

This method updates the properties based on the current state of the corresponding cloud object. This is primarily used for manual polling of compute state.

serialize

Convert this RemoteCompute object into a JSON serialized dictionary.

attach

DEPRECATED. Use the attach_configuration method instead.

Associate an existing remote compute resource with the provided workspace.

static attach(workspace, name, username, address, ssh_port=22, password='', private_key_file='', private_key_passphrase='')

Parameters

Name Description
workspace
Required

The workspace object to associate the compute resource with.

name
Required
str

The name to associate with the compute resource inside the provided workspace. Does not have to match the name of the compute resource to be attached.

username
Required
str

The username needed to access the resource.

address
Required
str

The address of the resource to be attached.

ssh_port
int

The exposed port for the resource. Defaults to 22.

Default value: 22
password
Required
str

The password needed to access the resource.

private_key_file
Required
str

Path to a file containing the private key for the resource.

private_key_passphrase
Required
str

Private key phrase needed to access the resource.

Returns

Type Description

A RemoteCompute object representation of the compute object.

Exceptions

Type Description

attach_configuration

Create a configuration object for attaching a remote compute target.

Attaching a virtual machine using the public IP address of the VM is no longer supported. Instead, use the resourceId of the VM. The resourceId of the VM can be constructed using the following string format: "/subscriptions/<subscription_id>/resourceGroups/<resource_group>/ providers/Microsoft.Compute/virtualMachines/<vm_name>".

You can also use subscription_id, resource_group and vm_name without constructing resourceId. For more information, see https://aka.ms/azureml-compute-vm.

static attach_configuration(username, subscription_id=None, resource_group=None, vm_name=None, resource_id=None, address=None, ssh_port=22, password='', private_key_file='', private_key_passphrase='')

Parameters

Name Description
username
Required
str

The username needed to access the resource.

subscription_id
str

The Azure subscription ID in which virtual machine is located.

Default value: None
resource_group
str

The name of the resource group in which virtual machine is located.

Default value: None
vm_name
str

The virtual machine name.

Default value: None
resource_id
str

The Azure Resource Manager (ARM) resource ID for the existing resource.

Default value: None
address
str

The address for the existing resource.

Default value: None
ssh_port
int

The exposed port for the resource. Defaults to 22.

Default value: 22
password
Required
str

The password needed to access the resource.

private_key_file
Required
str

Path to a file containing the private key for the resource.

private_key_passphrase
Required
str

The private key phrase needed to access the resource.

Returns

Type Description

A configuration object to be used when attaching a Compute object.

Exceptions

Type Description

delete

Delete is not supported for a RemoteCompute object. Use detach instead.

delete()

Exceptions

Type Description

deserialize

Convert a JSON object into a RemoteCompute object.

static deserialize(workspace, object_dict)

Parameters

Name Description
workspace
Required

The workspace object the RemoteCompute object is associated with.

object_dict
Required

A JSON object to convert to a RemoteCompute object.

Returns

Type Description

The RemoteCompute representation of the provided JSON object.

Exceptions

Type Description

Remarks

Raises a ComputeTargetException if the provided workspace is not the workspace the Compute is associated with.

detach

Detach the RemoteCompute object from its associated workspace.

Underlying cloud objects are not deleted, only the association is removed.

detach()

Exceptions

Type Description

get_credentials

Retrieve the credentials for the RemoteCompute target.

get_credentials()

Returns

Type Description

The credentials for the RemoteCompute target.

Exceptions

Type Description

refresh_state

Perform an in-place update of the properties of the object.

This method updates the properties based on the current state of the corresponding cloud object. This is primarily used for manual polling of compute state.

refresh_state()

Exceptions

Type Description

serialize

Convert this RemoteCompute object into a JSON serialized dictionary.

serialize()

Returns

Type Description

The JSON representation of this RemoteCompute object.

Exceptions

Type Description