azure.mgmt.batchai.models module

class azure.mgmt.batchai.models.AppInsightsReference(*, component: azure.mgmt.batchai.models._models_py3.ResourceId, instrumentation_key: Optional[str] = None, instrumentation_key_secret_reference: Optional[KeyVaultSecretReference] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Azure Application Insights information for performance counters reporting.

All required parameters must be populated in order to send to Azure.

Parameters

  • component (ResourceId) – Required. Azure Application Insights component resource ID.

  • instrumentation_key (str) – Value of the Azure Application Insights instrumentation key.

  • instrumentation_key_secret_reference (KeyVaultSecretReference) – KeyVault Store and Secret which contains Azure Application Insights instrumentation key. One of instrumentationKey or instrumentationKeySecretReference must be specified.

class azure.mgmt.batchai.models.AutoScaleSettings(*, minimum_node_count: int, maximum_node_count: int, initial_node_count: Optional[int] = 0, **kwargs)[source]

Bases: msrest.serialization.Model

Auto-scale settings for the cluster. The system automatically scales the cluster up and down (within minimumNodeCount and maximumNodeCount) based on the number of queued and running jobs assigned to the cluster.

All required parameters must be populated in order to send to Azure.

Parameters

  • minimum_node_count (int) – Required. The minimum number of compute nodes the Batch AI service will try to allocate for the cluster. Note, the actual number of nodes can be less than the specified value if the subscription has not enough quota to fulfill the request.

  • maximum_node_count (int) – Required. The maximum number of compute nodes the cluster can have.

  • initial_node_count (int) – The number of compute nodes to allocate on cluster creation. Note that this value is used only during cluster creation. Default: 0.

class azure.mgmt.batchai.models.AzureBlobFileSystemReference(*, account_name: str, container_name: str, credentials: azure.mgmt.batchai.models._models_py3.AzureStorageCredentialsInfo, relative_mount_path: str, mount_options: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Azure Blob Storage Container mounting configuration.

All required parameters must be populated in order to send to Azure.

Parameters

  • account_name (str) – Required. Name of the Azure storage account.

  • container_name (str) – Required. Name of the Azure Blob Storage container to mount on the cluster.

  • credentials (AzureStorageCredentialsInfo) – Required. Information about the Azure storage credentials.

  • relative_mount_path (str) – Required. The relative path on the compute node where the Azure File container will be mounted. Note that all cluster level containers will be mounted under $AZ_BATCHAI_MOUNT_ROOT location and all job level containers will be mounted under $AZ_BATCHAI_JOB_MOUNT_ROOT.

  • mount_options (str) – Mount options for mounting blobfuse file system.

class azure.mgmt.batchai.models.AzureFileShareReference(*, account_name: str, azure_file_url: str, credentials: azure.mgmt.batchai.models._models_py3.AzureStorageCredentialsInfo, relative_mount_path: str, file_mode: Optional[str] = '0777', directory_mode: Optional[str] = '0777', **kwargs)[source]

Bases: msrest.serialization.Model

Azure File Share mounting configuration.

All required parameters must be populated in order to send to Azure.

Parameters

  • account_name (str) – Required. Name of the Azure storage account.

  • azure_file_url (str) – Required. URL to access the Azure File.

  • credentials (AzureStorageCredentialsInfo) – Required. Information about the Azure storage credentials.

  • relative_mount_path (str) – Required. The relative path on the compute node where the Azure File share will be mounted. Note that all cluster level file shares will be mounted under $AZ_BATCHAI_MOUNT_ROOT location and all job level file shares will be mounted under $AZ_BATCHAI_JOB_MOUNT_ROOT.

  • file_mode (str) – File mode for files on the mounted file share. Default value: 0777.

  • directory_mode (str) – File mode for directories on the mounted file share. Default value: 0777.

class azure.mgmt.batchai.models.AzureStorageCredentialsInfo(*, account_key: Optional[str] = None, account_key_secret_reference: Optional[KeyVaultSecretReference] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Azure storage account credentials.

Parameters

  • account_key (str) – Storage account key. One of accountKey or accountKeySecretReference must be specified.

  • account_key_secret_reference (KeyVaultSecretReference) – Information about KeyVault secret storing the storage account key. One of accountKey or accountKeySecretReference must be specified.

class azure.mgmt.batchai.models.BatchAIError(**kwargs)[source]

Bases: msrest.serialization.Model

An error response from the Batch AI service.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • code (str) – An identifier of the error. Codes are invariant and are intended to be consumed programmatically.

  • message (str) – A message describing the error, intended to be suitable for display in a user interface.

  • details (list[NameValuePair]) – A list of additional details about the error.

class azure.mgmt.batchai.models.CNTKsettings(*, language_type: Optional[str] = None, config_file_path: Optional[str] = None, python_script_file_path: Optional[str] = None, python_interpreter_path: Optional[str] = None, command_line_args: Optional[str] = None, process_count: Optional[int] = None, **kwargs)[source]

Bases: msrest.serialization.Model

CNTK (aka Microsoft Cognitive Toolkit) job settings.

Parameters

  • language_type (str) – The language to use for launching CNTK (aka Microsoft Cognitive Toolkit) job. Valid values are ‘BrainScript’ or ‘Python’.

  • config_file_path (str) – Specifies the path of the BrainScript config file. This property can be specified only if the languageType is ‘BrainScript’.

  • python_script_file_path (str) – Python script to execute. This property can be specified only if the languageType is ‘Python’.

  • python_interpreter_path (str) – The path to the Python interpreter. This property can be specified only if the languageType is ‘Python’.

  • command_line_args (str) – Command line arguments that need to be passed to the python script or cntk executable.

  • process_count (int) – Number of processes to launch for the job execution. The default value for this property is equal to nodeCount property.

class azure.mgmt.batchai.models.Caffe2Settings(*, python_script_file_path: str, python_interpreter_path: Optional[str] = None, command_line_args: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Caffe2 job settings.

All required parameters must be populated in order to send to Azure.

Parameters

  • python_script_file_path (str) – Required. The python script to execute.

  • python_interpreter_path (str) – The path to the Python interpreter.

  • command_line_args (str) – Command line arguments that need to be passed to the python script.

class azure.mgmt.batchai.models.CaffeSettings(*, config_file_path: Optional[str] = None, python_script_file_path: Optional[str] = None, python_interpreter_path: Optional[str] = None, command_line_args: Optional[str] = None, process_count: Optional[int] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Caffe job settings.

Parameters

  • config_file_path (str) – Path of the config file for the job. This property cannot be specified if pythonScriptFilePath is specified.

  • python_script_file_path (str) – Python script to execute. This property cannot be specified if configFilePath is specified.

  • python_interpreter_path (str) – The path to the Python interpreter. The property can be specified only if the pythonScriptFilePath is specified.

  • command_line_args (str) – Command line arguments that need to be passed to the Caffe job.

  • process_count (int) – Number of processes to launch for the job execution. The default value for this property is equal to nodeCount property.

class azure.mgmt.batchai.models.ChainerSettings(*, python_script_file_path: str, python_interpreter_path: Optional[str] = None, command_line_args: Optional[str] = None, process_count: Optional[int] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Chainer job settings.

All required parameters must be populated in order to send to Azure.

Parameters

  • python_script_file_path (str) – Required. The python script to execute.

  • python_interpreter_path (str) – The path to the Python interpreter.

  • command_line_args (str) – Command line arguments that need to be passed to the python script.

  • process_count (int) – Number of processes to launch for the job execution. The default value for this property is equal to nodeCount property.

class azure.mgmt.batchai.models.CloudErrorBody(**kwargs)[source]

Bases: msrest.serialization.Model

An error response from the Batch AI service.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • code (str) – An identifier for the error. Codes are invariant and are intended to be consumed programmatically.

  • message (str) – A message describing the error, intended to be suitable for display in a user interface.

  • target (str) – The target of the particular error. For example, the name of the property in error.

  • details (list[CloudErrorBody]) – A list of additional details about the error.

class azure.mgmt.batchai.models.Cluster(*, vm_size: Optional[str] = None, vm_priority: Union[str, VmPriority, None] = None, scale_settings: Optional[ScaleSettings] = None, virtual_machine_configuration: Optional[VirtualMachineConfiguration] = None, node_setup: Optional[NodeSetup] = None, user_account_settings: Optional[UserAccountSettings] = None, subnet: Optional[ResourceId] = None, **kwargs)[source]

Bases: azure.mgmt.batchai.models._models_py3.ProxyResource

Information about a Cluster.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • id (str) – The ID of the resource.

  • name (str) – The name of the resource.

  • type (str) – The type of the resource.

  • creation_time (datetime) – The time when the cluster was created.

  • provisioning_state (str or ProvisioningState) – Provisioning state of the cluster. Possible value are: creating - Specifies that the cluster is being created. succeeded - Specifies that the cluster has been created successfully. failed - Specifies that the cluster creation has failed. deleting - Specifies that the cluster is being deleted. Possible values include: “creating”, “succeeded”, “failed”, “deleting”.

  • provisioning_state_transition_time (datetime) – Time when the provisioning state was changed.

  • allocation_state (str or AllocationState) – Allocation state of the cluster. Possible values are: steady - Indicates that the cluster is not resizing. There are no changes to the number of compute nodes in the cluster in progress. A cluster enters this state when it is created and when no operations are being performed on the cluster to change the number of compute nodes. resizing - Indicates that the cluster is resizing; that is, compute nodes are being added to or removed from the cluster. Possible values include: “steady”, “resizing”.

  • allocation_state_transition_time (datetime) – The time at which the cluster entered its current allocation state.

  • errors (list[BatchAIError]) – Collection of errors encountered by various compute nodes during node setup.

  • current_node_count (int) – The number of compute nodes currently assigned to the cluster.

  • node_state_counts (NodeStateCounts) – Counts of various node states on the cluster.

Parameters

  • vm_size (str) – The size of the virtual machines in the cluster. All nodes in a cluster have the same VM size.

  • vm_priority (str or VmPriority) – VM priority of cluster nodes. Possible values include: “dedicated”, “lowpriority”.

  • scale_settings (ScaleSettings) – Scale settings of the cluster.

  • virtual_machine_configuration (VirtualMachineConfiguration) – Virtual machine configuration (OS image) of the compute nodes. All nodes in a cluster have the same OS image configuration.

  • node_setup (NodeSetup) – Setup (mount file systems, performance counters settings and custom setup task) to be performed on each compute node in the cluster.

  • user_account_settings (UserAccountSettings) – Administrator user account settings which can be used to SSH to compute nodes.

  • subnet (ResourceId) – Virtual network subnet resource ID the cluster nodes belong to.

class azure.mgmt.batchai.models.ClusterCreateParameters(*, vm_size: Optional[str] = None, vm_priority: Union[str, VmPriority, None] = None, scale_settings: Optional[ScaleSettings] = None, virtual_machine_configuration: Optional[VirtualMachineConfiguration] = None, node_setup: Optional[NodeSetup] = None, user_account_settings: Optional[UserAccountSettings] = None, subnet: Optional[ResourceId] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Cluster creation operation.

Parameters

  • vm_size (str) – The size of the virtual machines in the cluster. All nodes in a cluster have the same VM size. For information about available VM sizes for clusters using images from the Virtual Machines Marketplace see Sizes for Virtual Machines (Linux). Batch AI service supports all Azure VM sizes except STANDARD_A0 and those with premium storage (STANDARD_GS, STANDARD_DS, and STANDARD_DSV2 series).

  • vm_priority (str or VmPriority) – VM priority. Allowed values are: dedicated (default) and lowpriority. Possible values include: “dedicated”, “lowpriority”.

  • scale_settings (ScaleSettings) – Scale settings for the cluster. Batch AI service supports manual and auto scale clusters.

  • virtual_machine_configuration (VirtualMachineConfiguration) – OS image configuration for cluster nodes. All nodes in a cluster have the same OS image.

  • node_setup (NodeSetup) – Setup to be performed on each compute node in the cluster.

  • user_account_settings (UserAccountSettings) – Settings for an administrator user account that will be created on each compute node in the cluster.

  • subnet (ResourceId) – Existing virtual network subnet to put the cluster nodes in. Note, if a File Server mount configured in node setup, the File Server’s subnet will be used automatically.

class azure.mgmt.batchai.models.ClusterListResult(**kwargs)[source]

Bases: msrest.serialization.Model

Values returned by the List Clusters operation.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • value (list[Cluster]) – The collection of returned Clusters.

  • next_link (str) – The continuation token.

class azure.mgmt.batchai.models.ClusterUpdateParameters(*, scale_settings: Optional[ScaleSettings] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Cluster update parameters.

Parameters

scale_settings (ScaleSettings) – Desired scale settings for the cluster. Batch AI service supports manual and auto scale clusters.

class azure.mgmt.batchai.models.ClustersListByWorkspaceOptions(*, max_results: Optional[int] = 1000, **kwargs)[source]

Bases: msrest.serialization.Model

Parameter group.

Parameters

max_results (int) – The maximum number of items to return in the response. A maximum of 1000 files can be returned.

class azure.mgmt.batchai.models.ContainerSettings(*, image_source_registry: azure.mgmt.batchai.models._models_py3.ImageSourceRegistry, shm_size: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Docker container settings.

All required parameters must be populated in order to send to Azure.

Parameters

  • image_source_registry (ImageSourceRegistry) – Required. Information about docker image and docker registry to download the container from.

  • shm_size (str) – Size of /dev/shm. Please refer to docker documentation for supported argument formats.

class azure.mgmt.batchai.models.CustomMpiSettings(*, command_line: str, process_count: Optional[int] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Custom MPI job settings.

All required parameters must be populated in order to send to Azure.

Parameters

  • command_line (str) – Required. The command line to be executed by mpi runtime on each compute node.

  • process_count (int) – Number of processes to launch for the job execution. The default value for this property is equal to nodeCount property.

class azure.mgmt.batchai.models.CustomToolkitSettings(*, command_line: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Custom tool kit job settings.

Parameters

command_line (str) – The command line to execute on the master node.

class azure.mgmt.batchai.models.DataDisks(*, disk_size_in_gb: int, disk_count: int, storage_account_type: Union[str, StorageAccountType], caching_type: Union[str, CachingType, None] = 'none', **kwargs)[source]

Bases: msrest.serialization.Model

Data disks settings.

All required parameters must be populated in order to send to Azure.

Parameters

  • disk_size_in_gb (int) – Required. Disk size in GB for the blank data disks.

  • caching_type (str or CachingType) – Caching type for the disks. Available values are none (default), readonly, readwrite. Caching type can be set only for VM sizes supporting premium storage. Possible values include: “none”, “readonly”, “readwrite”. Default value: “none”.

  • disk_count (int) – Required. Number of data disks attached to the File Server. If multiple disks attached, they will be configured in RAID level 0.

  • storage_account_type (str or StorageAccountType) – Required. Type of storage account to be used on the disk. Possible values are: Standard_LRS or Premium_LRS. Premium storage account type can only be used with VM sizes supporting premium storage. Possible values include: “Standard_LRS”, “Premium_LRS”.

class azure.mgmt.batchai.models.EnvironmentVariable(*, name: str, value: str, **kwargs)[source]

Bases: msrest.serialization.Model

An environment variable definition.

All required parameters must be populated in order to send to Azure.

Parameters

  • name (str) – Required. The name of the environment variable.

  • value (str) – Required. The value of the environment variable.

class azure.mgmt.batchai.models.EnvironmentVariableWithSecretValue(*, name: str, value: Optional[str] = None, value_secret_reference: Optional[KeyVaultSecretReference] = None, **kwargs)[source]

Bases: msrest.serialization.Model

An environment variable with secret value definition.

All required parameters must be populated in order to send to Azure.

Parameters

  • name (str) – Required. The name of the environment variable to store the secret value.

  • value (str) – The value of the environment variable. This value will never be reported back by Batch AI.

  • value_secret_reference (KeyVaultSecretReference) – KeyVault store and secret which contains the value for the environment variable. One of value or valueSecretReference must be provided.

class azure.mgmt.batchai.models.Experiment(**kwargs)[source]

Bases: azure.mgmt.batchai.models._models_py3.ProxyResource

Experiment information.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • id (str) – The ID of the resource.

  • name (str) – The name of the resource.

  • type (str) – The type of the resource.

  • creation_time (datetime) – Time when the Experiment was created.

  • provisioning_state (str or ProvisioningState) – The provisioned state of the experiment. Possible values include: “creating”, “succeeded”, “failed”, “deleting”.

  • provisioning_state_transition_time (datetime) – The time at which the experiment entered its current provisioning state.

class azure.mgmt.batchai.models.ExperimentListResult(**kwargs)[source]

Bases: msrest.serialization.Model

Values returned by the List operation.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • value (list[Experiment]) – The collection of experiments.

  • next_link (str) – The continuation token.

class azure.mgmt.batchai.models.ExperimentsListByWorkspaceOptions(*, max_results: Optional[int] = 1000, **kwargs)[source]

Bases: msrest.serialization.Model

Parameter group.

Parameters

max_results (int) – The maximum number of items to return in the response. A maximum of 1000 files can be returned.

class azure.mgmt.batchai.models.File(**kwargs)[source]

Bases: msrest.serialization.Model

Properties of the file or directory.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • name (str) – Name of the file.

  • file_type (str or FileType) – Type of the file. Possible values are file and directory. Possible values include: “file”, “directory”.

  • download_url (str) – URL to download the corresponding file. The downloadUrl is not returned for directories.

  • last_modified (datetime) – The time at which the file was last modified.

  • content_length (long) – The file of the size.

class azure.mgmt.batchai.models.FileListResult(**kwargs)[source]

Bases: msrest.serialization.Model

Values returned by the List operation.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • value (list[File]) – The collection of returned job directories and files.

  • next_link (str) – The continuation token.

class azure.mgmt.batchai.models.FileServer(*, vm_size: Optional[str] = None, ssh_configuration: Optional[SshConfiguration] = None, data_disks: Optional[DataDisks] = None, subnet: Optional[ResourceId] = None, **kwargs)[source]

Bases: azure.mgmt.batchai.models._models_py3.ProxyResource

File Server information.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • id (str) – The ID of the resource.

  • name (str) – The name of the resource.

  • type (str) – The type of the resource.

  • mount_settings (MountSettings) – File Server mount settings.

  • provisioning_state_transition_time (datetime) – Time when the provisioning state was changed.

  • creation_time (datetime) – Time when the FileServer was created.

  • provisioning_state (str or FileServerProvisioningState) – Provisioning state of the File Server. Possible values: creating - The File Server is getting created; updating - The File Server creation has been accepted and it is getting updated; deleting - The user has requested that the File Server be deleted, and it is in the process of being deleted; failed - The File Server creation has failed with the specified error code. Details about the error code are specified in the message field; succeeded - The File Server creation has succeeded. Possible values include: “creating”, “updating”, “deleting”, “succeeded”, “failed”.

Parameters

  • vm_size (str) – VM size of the File Server.

  • ssh_configuration (SshConfiguration) – SSH configuration for accessing the File Server node.

  • data_disks (DataDisks) – Information about disks attached to File Server VM.

  • subnet (ResourceId) – File Server virtual network subnet resource ID.

class azure.mgmt.batchai.models.FileServerCreateParameters(*, vm_size: Optional[str] = None, ssh_configuration: Optional[SshConfiguration] = None, data_disks: Optional[DataDisks] = None, subnet: Optional[ResourceId] = None, **kwargs)[source]

Bases: msrest.serialization.Model

File Server creation parameters.

Parameters

  • vm_size (str) – The size of the virtual machine for the File Server. For information about available VM sizes from the Virtual Machines Marketplace, see Sizes for Virtual Machines (Linux).

  • ssh_configuration (SshConfiguration) – SSH configuration for the File Server node.

  • data_disks (DataDisks) – Settings for the data disks which will be created for the File Server.

  • subnet (ResourceId) – Identifier of an existing virtual network subnet to put the File Server in. If not provided, a new virtual network and subnet will be created.

class azure.mgmt.batchai.models.FileServerListResult(*, value: Optional[List[FileServer]] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Values returned by the File Server List operation.

Variables are only populated by the server, and will be ignored when sending a request.

Parameters

value (list[FileServer]) – The collection of File Servers.

Variables

next_link (str) – The continuation token.

class azure.mgmt.batchai.models.FileServerReference(*, file_server: azure.mgmt.batchai.models._models_py3.ResourceId, relative_mount_path: str, source_directory: Optional[str] = None, mount_options: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

File Server mounting configuration.

All required parameters must be populated in order to send to Azure.

Parameters

  • file_server (ResourceId) – Required. Resource ID of the existing File Server to be mounted.

  • source_directory (str) – File Server directory that needs to be mounted. If this property is not specified, the entire File Server will be mounted.

  • relative_mount_path (str) – Required. The relative path on the compute node where the File Server will be mounted. Note that all cluster level file servers will be mounted under $AZ_BATCHAI_MOUNT_ROOT location and all job level file servers will be mounted under $AZ_BATCHAI_JOB_MOUNT_ROOT.

  • mount_options (str) – Mount options to be passed to mount command.

class azure.mgmt.batchai.models.FileServersListByWorkspaceOptions(*, max_results: Optional[int] = 1000, **kwargs)[source]

Bases: msrest.serialization.Model

Parameter group.

Parameters

max_results (int) – The maximum number of items to return in the response. A maximum of 1000 files can be returned.

class azure.mgmt.batchai.models.HorovodSettings(*, python_script_file_path: str, python_interpreter_path: Optional[str] = None, command_line_args: Optional[str] = None, process_count: Optional[int] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Specifies the settings for Horovod job.

All required parameters must be populated in order to send to Azure.

Parameters

  • python_script_file_path (str) – Required. The python script to execute.

  • python_interpreter_path (str) – The path to the Python interpreter.

  • command_line_args (str) – Command line arguments that need to be passed to the python script.

  • process_count (int) – Number of processes to launch for the job execution. The default value for this property is equal to nodeCount property.

class azure.mgmt.batchai.models.ImageReference(*, publisher: str, offer: str, sku: str, version: Optional[str] = None, virtual_machine_image_id: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

The OS image reference.

All required parameters must be populated in order to send to Azure.

Parameters

  • publisher (str) – Required. Publisher of the image.

  • offer (str) – Required. Offer of the image.

  • sku (str) – Required. SKU of the image.

  • version (str) – Version of the image.

  • virtual_machine_image_id (str) – The ARM resource identifier of the virtual machine image for the compute nodes. This is of the form /subscriptions/{subscriptionId}/resourceGroups/{resourceGroup}/providers/Microsoft.Compute/images/{imageName}. The virtual machine image must be in the same region and subscription as the cluster. For information about the firewall settings for the Batch node agent to communicate with the Batch service see https://docs.microsoft.com/en-us/azure/batch/batch-api-basics#virtual-network-vnet-and-firewall-configuration. Note, you need to provide publisher, offer and sku of the base OS image of which the custom image has been derived from.

class azure.mgmt.batchai.models.ImageSourceRegistry(*, image: str, server_url: Optional[str] = None, credentials: Optional[PrivateRegistryCredentials] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Information about docker image for the job.

All required parameters must be populated in order to send to Azure.

Parameters

  • server_url (str) – URL for image repository.

  • image (str) – Required. The name of the image in the image repository.

  • credentials (PrivateRegistryCredentials) – Credentials to access the private docker repository.

class azure.mgmt.batchai.models.InputDirectory(*, id: str, path: str, **kwargs)[source]

Bases: msrest.serialization.Model

Input directory for the job.

All required parameters must be populated in order to send to Azure.

Parameters

  • id (str) – Required. The ID for the input directory. The job can use AZ_BATCHAIINPUT <id> environment variable to find the directory path, where <id> is the value of id attribute.

  • path (str) – Required. The path to the input directory.

class azure.mgmt.batchai.models.Job(*, scheduling_priority: Union[str, JobPriority, None] = None, cluster: Optional[ResourceId] = None, mount_volumes: Optional[MountVolumes] = None, node_count: Optional[int] = None, container_settings: Optional[ContainerSettings] = None, tool_type: Union[str, ToolType, None] = None, cntk_settings: Optional[CNTKsettings] = None, py_torch_settings: Optional[PyTorchSettings] = None, tensor_flow_settings: Optional[TensorFlowSettings] = None, caffe_settings: Optional[CaffeSettings] = None, caffe2_settings: Optional[Caffe2Settings] = None, chainer_settings: Optional[ChainerSettings] = None, custom_toolkit_settings: Optional[CustomToolkitSettings] = None, custom_mpi_settings: Optional[CustomMpiSettings] = None, horovod_settings: Optional[HorovodSettings] = None, job_preparation: Optional[JobPreparation] = None, std_out_err_path_prefix: Optional[str] = None, input_directories: Optional[List[InputDirectory]] = None, output_directories: Optional[List[OutputDirectory]] = None, environment_variables: Optional[List[EnvironmentVariable]] = None, secrets: Optional[List[EnvironmentVariableWithSecretValue]] = None, constraints: Optional[JobPropertiesConstraints] = None, execution_info: Optional[JobPropertiesExecutionInfo] = None, **kwargs)[source]

Bases: azure.mgmt.batchai.models._models_py3.ProxyResource

Information about a Job.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • id (str) – The ID of the resource.

  • name (str) – The name of the resource.

  • type (str) – The type of the resource.

  • job_output_directory_path_segment (str) – A segment of job’s output directories path created by Batch AI. Batch AI creates job’s output directories under an unique path to avoid conflicts between jobs. This value contains a path segment generated by Batch AI to make the path unique and can be used to find the output directory on the node or mounted filesystem.

  • creation_time (datetime) – The creation time of the job.

  • provisioning_state (str or ProvisioningState) – The provisioned state of the Batch AI job. Possible values include: “creating”, “succeeded”, “failed”, “deleting”.

  • provisioning_state_transition_time (datetime) – The time at which the job entered its current provisioning state.

  • execution_state (str or ExecutionState) – The current state of the job. Possible values are: queued - The job is queued and able to run. A job enters this state when it is created, or when it is awaiting a retry after a failed run. running - The job is running on a compute cluster. This includes job-level preparation such as downloading resource files or set up container specified on the job - it does not necessarily mean that the job command line has started executing. terminating - The job is terminated by the user, the terminate operation is in progress. succeeded - The job has completed running successfully and exited with exit code 0. failed - The job has finished unsuccessfully (failed with a non-zero exit code) and has exhausted its retry limit. A job is also marked as failed if an error occurred launching the job. Possible values include: “queued”, “running”, “terminating”, “succeeded”, “failed”.

  • execution_state_transition_time (datetime) – The time at which the job entered its current execution state.

Parameters

  • scheduling_priority (str or JobPriority) – Scheduling priority associated with the job. Possible values include: “low”, “normal”, “high”.

  • cluster (ResourceId) – Resource ID of the cluster associated with the job.

  • mount_volumes (MountVolumes) – Collection of mount volumes available to the job during execution. These volumes are mounted before the job execution and unmounted after the job completion. The volumes are mounted at location specified by $AZ_BATCHAI_JOB_MOUNT_ROOT environment variable.

  • node_count (int) – The job will be gang scheduled on that many compute nodes.

  • container_settings (ContainerSettings) – If the container was downloaded as part of cluster setup then the same container image will be used. If not provided, the job will run on the VM.

  • tool_type (str or ToolType) – Possible values are: cntk, tensorflow, caffe, caffe2, chainer, pytorch, custom, custommpi, horovod. Possible values include: “cntk”, “tensorflow”, “caffe”, “caffe2”, “chainer”, “horovod”, “custommpi”, “custom”.

  • cntk_settings (CNTKsettings) – CNTK (aka Microsoft Cognitive Toolkit) job settings.

  • py_torch_settings (PyTorchSettings) – pyTorch job settings.

  • tensor_flow_settings (TensorFlowSettings) – TensorFlow job settings.

  • caffe_settings (CaffeSettings) – Caffe job settings.

  • caffe2_settings (Caffe2Settings) – Caffe2 job settings.

  • chainer_settings (ChainerSettings) – Chainer job settings.

  • custom_toolkit_settings (CustomToolkitSettings) – Custom tool kit job settings.

  • custom_mpi_settings (CustomMpiSettings) – Custom MPI job settings.

  • horovod_settings (HorovodSettings) – Specifies the settings for Horovod job.

  • job_preparation (JobPreparation) – The specified actions will run on all the nodes that are part of the job.

  • std_out_err_path_prefix (str) – The path where the Batch AI service stores stdout, stderror and execution log of the job.

  • input_directories (list[InputDirectory]) – A list of input directories for the job.

  • output_directories (list[OutputDirectory]) – A list of output directories for the job.

  • environment_variables (list[EnvironmentVariable]) – A collection of user defined environment variables to be setup for the job.

  • secrets (list[EnvironmentVariableWithSecretValue]) – A collection of user defined environment variables with secret values to be setup for the job. Server will never report values of these variables back.

  • constraints (JobPropertiesConstraints) – Constraints associated with the Job.

  • execution_info (JobPropertiesExecutionInfo) – Information about the execution of a job.

class azure.mgmt.batchai.models.JobBasePropertiesConstraints(*, max_wall_clock_time: Optional[datetime.timedelta] = '7.00:00:00', **kwargs)[source]

Bases: msrest.serialization.Model

Constraints associated with the Job.

Parameters

max_wall_clock_time (timedelta) – Max time the job can run. Default value: 1 week.

class azure.mgmt.batchai.models.JobCreateParameters(*, scheduling_priority: Union[str, JobPriority, None] = None, cluster: Optional[ResourceId] = None, mount_volumes: Optional[MountVolumes] = None, node_count: Optional[int] = None, container_settings: Optional[ContainerSettings] = None, cntk_settings: Optional[CNTKsettings] = None, py_torch_settings: Optional[PyTorchSettings] = None, tensor_flow_settings: Optional[TensorFlowSettings] = None, caffe_settings: Optional[CaffeSettings] = None, caffe2_settings: Optional[Caffe2Settings] = None, chainer_settings: Optional[ChainerSettings] = None, custom_toolkit_settings: Optional[CustomToolkitSettings] = None, custom_mpi_settings: Optional[CustomMpiSettings] = None, horovod_settings: Optional[HorovodSettings] = None, job_preparation: Optional[JobPreparation] = None, std_out_err_path_prefix: Optional[str] = None, input_directories: Optional[List[InputDirectory]] = None, output_directories: Optional[List[OutputDirectory]] = None, environment_variables: Optional[List[EnvironmentVariable]] = None, secrets: Optional[List[EnvironmentVariableWithSecretValue]] = None, constraints: Optional[JobBasePropertiesConstraints] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Job creation parameters.

Parameters

  • scheduling_priority (str or JobPriority) – Scheduling priority associated with the job. Possible values: low, normal, high. Possible values include: “low”, “normal”, “high”.

  • cluster (ResourceId) – Resource ID of the cluster on which this job will run.

  • mount_volumes (MountVolumes) – Information on mount volumes to be used by the job. These volumes will be mounted before the job execution and will be unmounted after the job completion. The volumes will be mounted at location specified by $AZ_BATCHAI_JOB_MOUNT_ROOT environment variable.

  • node_count (int) – Number of compute nodes to run the job on. The job will be gang scheduled on that many compute nodes.

  • container_settings (ContainerSettings) – Docker container settings for the job. If not provided, the job will run directly on the node.

  • cntk_settings (CNTKsettings) – Settings for CNTK (aka Microsoft Cognitive Toolkit) job.

  • py_torch_settings (PyTorchSettings) – Settings for pyTorch job.

  • tensor_flow_settings (TensorFlowSettings) – Settings for Tensor Flow job.

  • caffe_settings (CaffeSettings) – Settings for Caffe job.

  • caffe2_settings (Caffe2Settings) – Settings for Caffe2 job.

  • chainer_settings (ChainerSettings) – Settings for Chainer job.

  • custom_toolkit_settings (CustomToolkitSettings) – Settings for custom tool kit job.

  • custom_mpi_settings (CustomMpiSettings) – Settings for custom MPI job.

  • horovod_settings (HorovodSettings) – Settings for Horovod job.

  • job_preparation (JobPreparation) – A command line to be executed on each node allocated for the job before tool kit is launched.

  • std_out_err_path_prefix (str) – The path where the Batch AI service will store stdout, stderror and execution log of the job.

  • input_directories (list[InputDirectory]) – A list of input directories for the job.

  • output_directories (list[OutputDirectory]) – A list of output directories for the job.

  • environment_variables (list[EnvironmentVariable]) – A list of user defined environment variables which will be setup for the job.

  • secrets (list[EnvironmentVariableWithSecretValue]) – A list of user defined environment variables with secret values which will be setup for the job. Server will never report values of these variables back.

  • constraints (JobBasePropertiesConstraints) – Constraints associated with the Job.

class azure.mgmt.batchai.models.JobListResult(**kwargs)[source]

Bases: msrest.serialization.Model

Values returned by the List operation.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • value (list[Job]) – The collection of jobs.

  • next_link (str) – The continuation token.

class azure.mgmt.batchai.models.JobPreparation(*, command_line: str, **kwargs)[source]

Bases: msrest.serialization.Model

Job preparation settings.

All required parameters must be populated in order to send to Azure.

Parameters

command_line (str) – Required. The command line to execute. If containerSettings is specified on the job, this commandLine will be executed in the same container as job. Otherwise it will be executed on the node.

class azure.mgmt.batchai.models.JobPropertiesConstraints(*, max_wall_clock_time: Optional[datetime.timedelta] = '7.00:00:00', **kwargs)[source]

Bases: msrest.serialization.Model

Constraints associated with the Job.

Parameters

max_wall_clock_time (timedelta) – Max time the job can run. Default value: 1 week.

class azure.mgmt.batchai.models.JobPropertiesExecutionInfo(**kwargs)[source]

Bases: msrest.serialization.Model

Information about the execution of a job.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • start_time (datetime) – The time at which the job started running. ‘Running’ corresponds to the running state. If the job has been restarted or retried, this is the most recent time at which the job started running. This property is present only for job that are in the running or completed state.

  • end_time (datetime) – The time at which the job completed. This property is only returned if the job is in completed state.

  • exit_code (int) – The exit code of the job. This property is only returned if the job is in completed state.

  • errors (list[BatchAIError]) – A collection of errors encountered by the service during job execution.

class azure.mgmt.batchai.models.JobsListByExperimentOptions(*, max_results: Optional[int] = 1000, **kwargs)[source]

Bases: msrest.serialization.Model

Parameter group.

Parameters

max_results (int) – The maximum number of items to return in the response. A maximum of 1000 files can be returned.

class azure.mgmt.batchai.models.JobsListOutputFilesOptions(*, outputdirectoryid: str, directory: Optional[str] = '.', linkexpiryinminutes: Optional[int] = 60, max_results: Optional[int] = 1000, **kwargs)[source]

Bases: msrest.serialization.Model

Parameter group.

All required parameters must be populated in order to send to Azure.

Parameters

  • outputdirectoryid (str) – Required. Id of the job output directory. This is the OutputDirectory–>id parameter that is given by the user during Create Job.

  • directory (str) – The path to the directory.

  • linkexpiryinminutes (int) – The number of minutes after which the download link will expire.

  • max_results (int) – The maximum number of items to return in the response. A maximum of 1000 files can be returned.

class azure.mgmt.batchai.models.KeyVaultSecretReference(*, source_vault: azure.mgmt.batchai.models._models_py3.ResourceId, secret_url: str, **kwargs)[source]

Bases: msrest.serialization.Model

Key Vault Secret reference.

All required parameters must be populated in order to send to Azure.

Parameters

  • source_vault (ResourceId) – Required. Fully qualified resource identifier of the Key Vault.

  • secret_url (str) – Required. The URL referencing a secret in the Key Vault.

class azure.mgmt.batchai.models.ListUsagesResult(**kwargs)[source]

Bases: msrest.serialization.Model

The List Usages operation response.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • value (list[Usage]) – The list of compute resource usages.

  • next_link (str) – The URI to fetch the next page of compute resource usage information. Call ListNext() with this to fetch the next page of compute resource usage information.

class azure.mgmt.batchai.models.ManualScaleSettings(*, target_node_count: int, node_deallocation_option: Union[str, DeallocationOption, None] = 'requeue', **kwargs)[source]

Bases: msrest.serialization.Model

Manual scale settings for the cluster.

All required parameters must be populated in order to send to Azure.

Parameters

  • target_node_count (int) – Required. The desired number of compute nodes in the Cluster. Default is 0.

  • node_deallocation_option (str or DeallocationOption) – An action to be performed when the cluster size is decreasing. The default value is requeue. Possible values include: “requeue”, “terminate”, “waitforjobcompletion”. Default value: “requeue”.

class azure.mgmt.batchai.models.MountSettings(*, mount_point: Optional[str] = None, file_server_public_ip: Optional[str] = None, file_server_internal_ip: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

File Server mount Information.

Parameters

  • mount_point (str) – Path where the data disks are mounted on the File Server.

  • file_server_public_ip (str) – Public IP address of the File Server which can be used to SSH to the node from outside of the subnet.

  • file_server_internal_ip (str) – Internal IP address of the File Server which can be used to access the File Server from within the subnet.

class azure.mgmt.batchai.models.MountVolumes(*, azure_file_shares: Optional[List[AzureFileShareReference]] = None, azure_blob_file_systems: Optional[List[AzureBlobFileSystemReference]] = None, file_servers: Optional[List[FileServerReference]] = None, unmanaged_file_systems: Optional[List[UnmanagedFileSystemReference]] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Details of volumes to mount on the cluster.

Parameters

  • azure_file_shares (list[AzureFileShareReference]) – A collection of Azure File Shares that are to be mounted to the cluster nodes.

  • azure_blob_file_systems (list[AzureBlobFileSystemReference]) – A collection of Azure Blob Containers that are to be mounted to the cluster nodes.

  • file_servers (list[FileServerReference]) – A collection of Batch AI File Servers that are to be mounted to the cluster nodes.

  • unmanaged_file_systems (list[UnmanagedFileSystemReference]) – A collection of unmanaged file systems that are to be mounted to the cluster nodes.

class azure.mgmt.batchai.models.NameValuePair(*, name: Optional[str] = None, value: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Name-value pair.

Parameters

  • name (str) – The name in the name-value pair.

  • value (str) – The value in the name-value pair.

class azure.mgmt.batchai.models.NodeSetup(*, setup_task: Optional[SetupTask] = None, mount_volumes: Optional[MountVolumes] = None, performance_counters_settings: Optional[PerformanceCountersSettings] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Node setup settings.

Parameters

  • setup_task (SetupTask) – Setup task to run on cluster nodes when nodes got created or rebooted. The setup task code needs to be idempotent. Generally the setup task is used to download static data that is required for all jobs that run on the cluster VMs and/or to download/install software.

  • mount_volumes (MountVolumes) – Mount volumes to be available to setup task and all jobs executing on the cluster. The volumes will be mounted at location specified by $AZ_BATCHAI_MOUNT_ROOT environment variable.

  • performance_counters_settings (PerformanceCountersSettings) – Settings for performance counters collecting and uploading.

class azure.mgmt.batchai.models.NodeStateCounts(**kwargs)[source]

Bases: msrest.serialization.Model

Counts of various compute node states on the cluster.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • idle_node_count (int) – Number of compute nodes in idle state.

  • running_node_count (int) – Number of compute nodes which are running jobs.

  • preparing_node_count (int) – Number of compute nodes which are being prepared.

  • unusable_node_count (int) – Number of compute nodes which are in unusable state.

  • leaving_node_count (int) – Number of compute nodes which are leaving the cluster.

class azure.mgmt.batchai.models.Operation(*, display: Optional[OperationDisplay] = None, properties: Optional[Any] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Details of a REST API operation.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • name (str) – This is of the format {provider}/{resource}/{operation}.

  • origin (str) – The intended executor of the operation.

Parameters

  • display (OperationDisplay) – The object that describes the operation.

  • properties (any) – Any object.

class azure.mgmt.batchai.models.OperationDisplay(**kwargs)[source]

Bases: msrest.serialization.Model

The object that describes the operation.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • provider (str) – Friendly name of the resource provider.

  • operation (str) – For example: read, write, delete, or listKeys/action.

  • resource (str) – The resource type on which the operation is performed.

  • description (str) – The friendly name of the operation.

class azure.mgmt.batchai.models.OperationListResult(**kwargs)[source]

Bases: msrest.serialization.Model

Contains the list of all operations supported by BatchAI resource provider.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • value (list[Operation]) – The list of operations supported by the resource provider.

  • next_link (str) – The URL to get the next set of operation list results if there are any.

class azure.mgmt.batchai.models.OutputDirectory(*, id: str, path_prefix: str, path_suffix: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Output directory for the job.

All required parameters must be populated in order to send to Azure.

Parameters

  • id (str) – Required. The ID of the output directory. The job can use AZ_BATCHAIOUTPUT <id> environment variable to find the directory path, where <id> is the value of id attribute.

  • path_prefix (str) – Required. The prefix path where the output directory will be created. Note, this is an absolute path to prefix. E.g. $AZ_BATCHAI_MOUNT_ROOT/MyNFS/MyLogs. The full path to the output directory by combining pathPrefix, jobOutputDirectoryPathSegment (reported by get job) and pathSuffix.

  • path_suffix (str) – The suffix path where the output directory will be created. E.g. models. You can find the full path to the output directory by combining pathPrefix, jobOutputDirectoryPathSegment (reported by get job) and pathSuffix.

class azure.mgmt.batchai.models.PerformanceCountersSettings(*, app_insights_reference: azure.mgmt.batchai.models._models_py3.AppInsightsReference, **kwargs)[source]

Bases: msrest.serialization.Model

Performance counters reporting settings.

All required parameters must be populated in order to send to Azure.

Parameters

app_insights_reference (AppInsightsReference) – Required. Azure Application Insights information for performance counters reporting. If provided, Batch AI will upload node performance counters to the corresponding Azure Application Insights account.

class azure.mgmt.batchai.models.PrivateRegistryCredentials(*, username: str, password: Optional[str] = None, password_secret_reference: Optional[KeyVaultSecretReference] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Credentials to access a container image in a private repository.

All required parameters must be populated in order to send to Azure.

Parameters

  • username (str) – Required. User name to login to the repository.

  • password (str) – User password to login to the docker repository. One of password or passwordSecretReference must be specified.

  • password_secret_reference (KeyVaultSecretReference) – KeyVault Secret storing the password. Users can store their secrets in Azure KeyVault and pass it to the Batch AI service to integrate with KeyVault. One of password or passwordSecretReference must be specified.

class azure.mgmt.batchai.models.ProxyResource(**kwargs)[source]

Bases: msrest.serialization.Model

A definition of an Azure proxy resource.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • id (str) – The ID of the resource.

  • name (str) – The name of the resource.

  • type (str) – The type of the resource.

class azure.mgmt.batchai.models.PyTorchSettings(*, python_script_file_path: str, python_interpreter_path: Optional[str] = None, command_line_args: Optional[str] = None, process_count: Optional[int] = None, communication_backend: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

pyTorch job settings.

All required parameters must be populated in order to send to Azure.

Parameters

  • python_script_file_path (str) – Required. The python script to execute.

  • python_interpreter_path (str) – The path to the Python interpreter.

  • command_line_args (str) – Command line arguments that need to be passed to the python script.

  • process_count (int) – Number of processes to launch for the job execution. The default value for this property is equal to nodeCount property.

  • communication_backend (str) – Type of the communication backend for distributed jobs. Valid values are ‘TCP’, ‘Gloo’ or ‘MPI’. Not required for non-distributed jobs.

class azure.mgmt.batchai.models.RemoteLoginInformation(**kwargs)[source]

Bases: msrest.serialization.Model

Login details to SSH to a compute node in cluster.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • node_id (str) – ID of the compute node.

  • ip_address (str) – Public IP address of the compute node.

  • port (int) – SSH port number of the node.

class azure.mgmt.batchai.models.RemoteLoginInformationListResult(**kwargs)[source]

Bases: msrest.serialization.Model

Values returned by the List operation.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • value (list[RemoteLoginInformation]) – The collection of returned remote login details.

  • next_link (str) – The continuation token.

class azure.mgmt.batchai.models.Resource(**kwargs)[source]

Bases: msrest.serialization.Model

A definition of an Azure resource.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • id (str) – The ID of the resource.

  • name (str) – The name of the resource.

  • type (str) – The type of the resource.

  • location (str) – The location of the resource.

  • tags (dict[str, str]) – A set of tags. The tags of the resource.

class azure.mgmt.batchai.models.ResourceId(*, id: str, **kwargs)[source]

Bases: msrest.serialization.Model

Represents a resource ID. For example, for a subnet, it is the resource URL for the subnet.

All required parameters must be populated in order to send to Azure.

Parameters

id (str) – Required. The ID of the resource.

class azure.mgmt.batchai.models.ScaleSettings(*, manual: Optional[ManualScaleSettings] = None, auto_scale: Optional[AutoScaleSettings] = None, **kwargs)[source]

Bases: msrest.serialization.Model

At least one of manual or autoScale settings must be specified. Only one of manual or autoScale settings can be specified. If autoScale settings are specified, the system automatically scales the cluster up and down (within the supplied limits) based on the pending jobs on the cluster.

Parameters

  • manual (ManualScaleSettings) – Manual scale settings for the cluster.

  • auto_scale (AutoScaleSettings) – Auto-scale settings for the cluster.

class azure.mgmt.batchai.models.SetupTask(*, command_line: str, std_out_err_path_prefix: str, environment_variables: Optional[List[EnvironmentVariable]] = None, secrets: Optional[List[EnvironmentVariableWithSecretValue]] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Specifies a setup task which can be used to customize the compute nodes of the cluster.

Variables are only populated by the server, and will be ignored when sending a request.

All required parameters must be populated in order to send to Azure.

Parameters

  • command_line (str) – Required. The command line to be executed on each cluster’s node after it being allocated or rebooted. The command is executed in a bash subshell as a root.

  • environment_variables (list[EnvironmentVariable]) – A collection of user defined environment variables to be set for setup task.

  • secrets (list[EnvironmentVariableWithSecretValue]) – A collection of user defined environment variables with secret values to be set for the setup task. Server will never report values of these variables back.

  • std_out_err_path_prefix (str) – Required. The prefix of a path where the Batch AI service will upload the stdout, stderr and execution log of the setup task.

Variables

std_out_err_path_suffix (str) – A path segment appended by Batch AI to stdOutErrPathPrefix to form a path where stdout, stderr and execution log of the setup task will be uploaded. Batch AI creates the setup task output directories under an unique path to avoid conflicts between different clusters. The full path can be obtained by concatenation of stdOutErrPathPrefix and stdOutErrPathSuffix.

class azure.mgmt.batchai.models.SshConfiguration(*, user_account_settings: azure.mgmt.batchai.models._models_py3.UserAccountSettings, public_ips_to_allow: Optional[List[str]] = None, **kwargs)[source]

Bases: msrest.serialization.Model

SSH configuration.

All required parameters must be populated in order to send to Azure.

Parameters

  • public_ips_to_allow (list[str]) – List of source IP ranges to allow SSH connection from. The default value is ‘*’ (all source IPs are allowed). Maximum number of IP ranges that can be specified is 400.

  • user_account_settings (UserAccountSettings) – Required. Settings for administrator user account to be created on a node. The account can be used to establish SSH connection to the node.

class azure.mgmt.batchai.models.TensorFlowSettings(*, python_script_file_path: str, python_interpreter_path: Optional[str] = None, master_command_line_args: Optional[str] = None, worker_command_line_args: Optional[str] = None, parameter_server_command_line_args: Optional[str] = None, worker_count: Optional[int] = None, parameter_server_count: Optional[int] = None, **kwargs)[source]

Bases: msrest.serialization.Model

TensorFlow job settings.

All required parameters must be populated in order to send to Azure.

Parameters

  • python_script_file_path (str) – Required. The python script to execute.

  • python_interpreter_path (str) – The path to the Python interpreter.

  • master_command_line_args (str) – Command line arguments that need to be passed to the python script for the master task.

  • worker_command_line_args (str) – Command line arguments that need to be passed to the python script for the worker task. Optional for single process jobs.

  • parameter_server_command_line_args (str) – Command line arguments that need to be passed to the python script for the parameter server. Optional for single process jobs.

  • worker_count (int) – The number of worker tasks. If specified, the value must be less than or equal to (nodeCount * numberOfGPUs per VM). If not specified, the default value is equal to nodeCount. This property can be specified only for distributed TensorFlow training.

  • parameter_server_count (int) – The number of parameter server tasks. If specified, the value must be less than or equal to nodeCount. If not specified, the default value is equal to 1 for distributed TensorFlow training. This property can be specified only for distributed TensorFlow training.

class azure.mgmt.batchai.models.UnmanagedFileSystemReference(*, mount_command: str, relative_mount_path: str, **kwargs)[source]

Bases: msrest.serialization.Model

Unmanaged file system mounting configuration.

All required parameters must be populated in order to send to Azure.

Parameters

  • mount_command (str) – Required. Mount command line. Note, Batch AI will append mount path to the command on its own.

  • relative_mount_path (str) – Required. The relative path on the compute node where the unmanaged file system will be mounted. Note that all cluster level unmanaged file systems will be mounted under $AZ_BATCHAI_MOUNT_ROOT location and all job level unmanaged file systems will be mounted under $AZ_BATCHAI_JOB_MOUNT_ROOT.

class azure.mgmt.batchai.models.Usage(**kwargs)[source]

Bases: msrest.serialization.Model

Describes Batch AI Resource Usage.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • unit (str or UsageUnit) – An enum describing the unit of usage measurement. Possible values include: “Count”.

  • current_value (int) – The current usage of the resource.

  • limit (long) – The maximum permitted usage of the resource.

  • name (UsageName) – The name of the type of usage.

class azure.mgmt.batchai.models.UsageName(**kwargs)[source]

Bases: msrest.serialization.Model

The Usage Names.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • value (str) – The name of the resource.

  • localized_value (str) – The localized name of the resource.

class azure.mgmt.batchai.models.UserAccountSettings(*, admin_user_name: str, admin_user_ssh_public_key: Optional[str] = None, admin_user_password: Optional[str] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Settings for user account that gets created on each on the nodes of a cluster.

All required parameters must be populated in order to send to Azure.

Parameters

  • admin_user_name (str) – Required. Name of the administrator user account which can be used to SSH to nodes.

  • admin_user_ssh_public_key (str) – SSH public key of the administrator user account.

  • admin_user_password (str) – Password of the administrator user account.

class azure.mgmt.batchai.models.VirtualMachineConfiguration(*, image_reference: Optional[ImageReference] = None, **kwargs)[source]

Bases: msrest.serialization.Model

VM configuration.

Parameters

image_reference (ImageReference) – OS image reference for cluster nodes.

class azure.mgmt.batchai.models.Workspace(**kwargs)[source]

Bases: azure.mgmt.batchai.models._models_py3.Resource

Batch AI Workspace information.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • id (str) – The ID of the resource.

  • name (str) – The name of the resource.

  • type (str) – The type of the resource.

  • location (str) – The location of the resource.

  • tags (dict[str, str]) – A set of tags. The tags of the resource.

  • creation_time (datetime) – Time when the Workspace was created.

  • provisioning_state (str or ProvisioningState) – The provisioned state of the Workspace. Possible values include: “creating”, “succeeded”, “failed”, “deleting”.

  • provisioning_state_transition_time (datetime) – The time at which the workspace entered its current provisioning state.

class azure.mgmt.batchai.models.WorkspaceCreateParameters(*, location: str, tags: Optional[Dict[str, str]] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Workspace creation parameters.

All required parameters must be populated in order to send to Azure.

Parameters

  • location (str) – Required. The region in which to create the Workspace.

  • tags (dict[str, str]) – A set of tags. The user specified tags associated with the Workspace.

class azure.mgmt.batchai.models.WorkspaceListResult(**kwargs)[source]

Bases: msrest.serialization.Model

Values returned by the List operation.

Variables are only populated by the server, and will be ignored when sending a request.

Variables

  • value (list[Workspace]) – The collection of workspaces.

  • next_link (str) – The continuation token.

class azure.mgmt.batchai.models.WorkspaceUpdateParameters(*, tags: Optional[Dict[str, str]] = None, **kwargs)[source]

Bases: msrest.serialization.Model

Workspace update parameters.

Parameters

tags (dict[str, str]) – A set of tags. The user specified tags associated with the Workspace.

class azure.mgmt.batchai.models.WorkspacesListByResourceGroupOptions(*, max_results: Optional[int] = 1000, **kwargs)[source]

Bases: msrest.serialization.Model

Parameter group.

Parameters

max_results (int) – The maximum number of items to return in the response. A maximum of 1000 files can be returned.

class azure.mgmt.batchai.models.WorkspacesListOptions(*, max_results: Optional[int] = 1000, **kwargs)[source]

Bases: msrest.serialization.Model

Parameter group.

Parameters

max_results (int) – The maximum number of items to return in the response. A maximum of 1000 files can be returned.

class azure.mgmt.batchai.models.AllocationState[source]

Bases: str, enum.Enum

Allocation state of the cluster. Possible values are: steady - Indicates that the cluster is not resizing. There are no changes to the number of compute nodes in the cluster in progress. A cluster enters this state when it is created and when no operations are being performed on the cluster to change the number of compute nodes. resizing - Indicates that the cluster is resizing; that is, compute nodes are being added to or removed from the cluster.

RESIZING = 'resizing'
STEADY = 'steady'
class azure.mgmt.batchai.models.CachingType[source]

Bases: str, enum.Enum

Caching type for the disks. Available values are none (default), readonly, readwrite. Caching type can be set only for VM sizes supporting premium storage.

NONE = 'none'
READONLY = 'readonly'
READWRITE = 'readwrite'
class azure.mgmt.batchai.models.DeallocationOption[source]

Bases: str, enum.Enum

Actions which should be performed when compute nodes are removed from the cluster. Possible values are: requeue (default) - Terminate running jobs and requeue them so the jobs will run again. Remove compute nodes as soon as jobs have been terminated; terminate - Terminate running jobs. The jobs will not run again. Remove compute nodes as soon as jobs have been terminated. waitforjobcompletion - Allow currently running jobs to complete. Schedule no new jobs while waiting. Remove compute nodes when all jobs have completed.

REQUEUE = 'requeue'
TERMINATE = 'terminate'
WAITFORJOBCOMPLETION = 'waitforjobcompletion'
class azure.mgmt.batchai.models.ExecutionState[source]

Bases: str, enum.Enum

The current state of the job. Possible values are: queued - The job is queued and able to run. A job enters this state when it is created, or when it is awaiting a retry after a failed run. running - The job is running on a compute cluster. This includes job-level preparation such as downloading resource files or set up container specified on the job - it does not necessarily mean that the job command line has started executing. terminating - The job is terminated by the user, the terminate operation is in progress. succeeded - The job has completed running successfully and exited with exit code 0. failed - The job has finished unsuccessfully (failed with a non-zero exit code) and has exhausted its retry limit. A job is also marked as failed if an error occurred launching the job.

FAILED = 'failed'
QUEUED = 'queued'
RUNNING = 'running'
SUCCEEDED = 'succeeded'
TERMINATING = 'terminating'
class azure.mgmt.batchai.models.FileServerProvisioningState[source]

Bases: str, enum.Enum

Provisioning state of the File Server. Possible values: creating - The File Server is getting created; updating - The File Server creation has been accepted and it is getting updated; deleting - The user has requested that the File Server be deleted, and it is in the process of being deleted; failed - The File Server creation has failed with the specified error code. Details about the error code are specified in the message field; succeeded - The File Server creation has succeeded.

CREATING = 'creating'
DELETING = 'deleting'
FAILED = 'failed'
SUCCEEDED = 'succeeded'
UPDATING = 'updating'
class azure.mgmt.batchai.models.FileType[source]

Bases: str, enum.Enum

Type of the file. Possible values are file and directory.

DIRECTORY = 'directory'
FILE = 'file'
class azure.mgmt.batchai.models.JobPriority[source]

Bases: str, enum.Enum

Scheduling priority associated with the job. Possible values: low, normal, high.

HIGH = 'high'
LOW = 'low'
NORMAL = 'normal'
class azure.mgmt.batchai.models.ProvisioningState[source]

Bases: str, enum.Enum

Provisioning state of the cluster. Possible value are: creating - Specifies that the cluster is being created. succeeded - Specifies that the cluster has been created successfully. failed - Specifies that the cluster creation has failed. deleting - Specifies that the cluster is being deleted.

CREATING = 'creating'
DELETING = 'deleting'
FAILED = 'failed'
SUCCEEDED = 'succeeded'
class azure.mgmt.batchai.models.StorageAccountType[source]

Bases: str, enum.Enum

Type of storage account to be used on the disk. Possible values are: Standard_LRS or Premium_LRS. Premium storage account type can only be used with VM sizes supporting premium storage.

PREMIUM_LRS = 'Premium_LRS'
STANDARD_LRS = 'Standard_LRS'
class azure.mgmt.batchai.models.ToolType[source]

Bases: str, enum.Enum

The toolkit type of the job.

CAFFE = 'caffe'
CAFFE2 = 'caffe2'
CHAINER = 'chainer'
CNTK = 'cntk'
CUSTOM = 'custom'
CUSTOMMPI = 'custommpi'
HOROVOD = 'horovod'
TENSORFLOW = 'tensorflow'
class azure.mgmt.batchai.models.UsageUnit[source]

Bases: str, enum.Enum

An enum describing the unit of usage measurement.

COUNT = 'Count'
class azure.mgmt.batchai.models.VmPriority[source]

Bases: str, enum.Enum

VM priority. Allowed values are: dedicated (default) and lowpriority.

DEDICATED = 'dedicated'
LOWPRIORITY = 'lowpriority'