AWS Parallel Computing Service

2025/08/01 - AWS Parallel Computing Service - 3 updated api methods

Changes  Add support for IPv6 Networking for Clusters.

CreateCluster (updated) Link ¶
Changes (request, response)
Request
{'networking': {'networkType': 'IPV4 | IPV6'}}
Response
{'cluster': {'endpoints': {'ipv6Address': 'string'},
             'networking': {'networkType': 'IPV4 | IPV6'}}}

Creates a cluster in your account. Amazon Web Services PCS creates the cluster controller in a service-owned account. The cluster controller communicates with the cluster resources in your account. The subnets and security groups for the cluster must already exist before you use this API action.

See also: AWS API Documentation

Request Syntax

client.create_cluster(
    clusterName='string',
    scheduler={
        'type': 'SLURM',
        'version': 'string'
    },
    size='SMALL'|'MEDIUM'|'LARGE',
    networking={
        'subnetIds': [
            'string',
        ],
        'securityGroupIds': [
            'string',
        ],
        'networkType': 'IPV4'|'IPV6'
    },
    slurmConfiguration={
        'scaleDownIdleTimeInSeconds': 123,
        'slurmCustomSettings': [
            {
                'parameterName': 'string',
                'parameterValue': 'string'
            },
        ],
        'accounting': {
            'mode': 'STANDARD'|'NONE',
            'defaultPurgeTimeInDays': 123
        }
    },
    clientToken='string',
    tags={
        'string': 'string'
    }
)
type clusterName:

string

param clusterName:

[REQUIRED]

A name to identify the cluster. Example: MyCluster

type scheduler:

dict

param scheduler:

[REQUIRED]

The cluster management and job scheduling software associated with the cluster.

  • type (string) -- [REQUIRED]

    The software Amazon Web Services PCS uses to manage cluster scaling and job scheduling.

  • version (string) -- [REQUIRED]

    The version of the specified scheduling software that Amazon Web Services PCS uses to manage cluster scaling and job scheduling. For more information, see Slurm versions in Amazon Web Services PCS in the Amazon Web Services PCS User Guide.

    Valid Values: 23.11 | 24.05 | 24.11

type size:

string

param size:

[REQUIRED]

A value that determines the maximum number of compute nodes in the cluster and the maximum number of jobs (active and queued).

  • SMALL: 32 compute nodes and 256 jobs

  • MEDIUM: 512 compute nodes and 8192 jobs

  • LARGE: 2048 compute nodes and 16,384 jobs

type networking:

dict

param networking:

[REQUIRED]

The networking configuration used to set up the cluster's control plane.

  • subnetIds (list) --

    The list of subnet IDs where Amazon Web Services PCS creates an Elastic Network Interface (ENI) to enable communication between managed controllers and Amazon Web Services PCS resources. Subnet IDs have the form subnet-0123456789abcdef0.

    Subnets can't be in Outposts, Wavelength or an Amazon Web Services Local Zone.

    • (string) --

  • securityGroupIds (list) --

    A list of security group IDs associated with the Elastic Network Interface (ENI) created in subnets.

    • (string) --

  • networkType (string) --

    The IP address version the cluster uses. The default is IPV4.

type slurmConfiguration:

dict

param slurmConfiguration:

Additional options related to the Slurm scheduler.

  • scaleDownIdleTimeInSeconds (integer) --

    The time (in seconds) before an idle node is scaled down.

    Default: 600

  • slurmCustomSettings (list) --

    Additional Slurm-specific configuration that directly maps to Slurm settings.

    • (dict) --

      Additional settings that directly map to Slurm settings.

      • parameterName (string) -- [REQUIRED]

        Amazon Web Services PCS supports configuration of the following Slurm parameters:

      • parameterValue (string) -- [REQUIRED]

        The values for the configured Slurm settings.

  • accounting (dict) --

    The accounting configuration includes configurable settings for Slurm accounting.

    • mode (string) -- [REQUIRED]

      The default value for mode is STANDARD. A value of STANDARD means Slurm accounting is enabled.

    • defaultPurgeTimeInDays (integer) --

      The default value for all purge settings for slurmdbd.conf. For more information, see the slurmdbd.conf documentation at SchedMD.

      The default value for defaultPurgeTimeInDays is -1.

      A value of -1 means there is no purge time and records persist as long as the cluster exists.

type clientToken:

string

param clientToken:

A unique, case-sensitive identifier that you provide to ensure the idempotency of the request. Idempotency ensures that an API request completes only once. With an idempotent request, if the original request completes successfully, the subsequent retries with the same client token return the result from the original successful request and they have no additional effect. If you don't specify a client token, the CLI and SDK automatically generate 1 for you.

This field is autopopulated if not provided.

type tags:

dict

param tags:

1 or more tags added to the resource. Each tag consists of a tag key and tag value. The tag value is optional and can be an empty string.

  • (string) --

    • (string) --

rtype:

dict

returns:

Response Syntax

{
    'cluster': {
        'name': 'string',
        'id': 'string',
        'arn': 'string',
        'status': 'CREATING'|'ACTIVE'|'UPDATING'|'DELETING'|'CREATE_FAILED'|'DELETE_FAILED'|'UPDATE_FAILED'|'SUSPENDING'|'SUSPENDED',
        'createdAt': datetime(2015, 1, 1),
        'modifiedAt': datetime(2015, 1, 1),
        'scheduler': {
            'type': 'SLURM',
            'version': 'string'
        },
        'size': 'SMALL'|'MEDIUM'|'LARGE',
        'slurmConfiguration': {
            'scaleDownIdleTimeInSeconds': 123,
            'slurmCustomSettings': [
                {
                    'parameterName': 'string',
                    'parameterValue': 'string'
                },
            ],
            'authKey': {
                'secretArn': 'string',
                'secretVersion': 'string'
            },
            'accounting': {
                'mode': 'STANDARD'|'NONE',
                'defaultPurgeTimeInDays': 123
            }
        },
        'networking': {
            'subnetIds': [
                'string',
            ],
            'securityGroupIds': [
                'string',
            ],
            'networkType': 'IPV4'|'IPV6'
        },
        'endpoints': [
            {
                'type': 'SLURMCTLD'|'SLURMDBD',
                'privateIpAddress': 'string',
                'publicIpAddress': 'string',
                'ipv6Address': 'string',
                'port': 'string'
            },
        ],
        'errorInfo': [
            {
                'code': 'string',
                'message': 'string'
            },
        ]
    }
}

Response Structure

  • (dict) --

    • cluster (dict) --

      The cluster resource.

      • name (string) --

        The name that identifies the cluster.

      • id (string) --

        The generated unique ID of the cluster.

      • arn (string) --

        The unique Amazon Resource Name (ARN) of the cluster.

      • status (string) --

        The provisioning status of the cluster.

      • createdAt (datetime) --

        The date and time the resource was created.

      • modifiedAt (datetime) --

        The date and time the resource was modified.

      • scheduler (dict) --

        The cluster management and job scheduling software associated with the cluster.

        • type (string) --

          The software Amazon Web Services PCS uses to manage cluster scaling and job scheduling.

        • version (string) --

          The version of the specified scheduling software that Amazon Web Services PCS uses to manage cluster scaling and job scheduling. For more information, see Slurm versions in Amazon Web Services PCS in the Amazon Web Services PCS User Guide.

          Valid Values: 23.11 | 24.05 | 24.11

      • size (string) --

        The size of the cluster.

        • SMALL: 32 compute nodes and 256 jobs

        • MEDIUM: 512 compute nodes and 8192 jobs

        • LARGE: 2048 compute nodes and 16,384 jobs

      • slurmConfiguration (dict) --

        Additional options related to the Slurm scheduler.

        • scaleDownIdleTimeInSeconds (integer) --

          The time (in seconds) before an idle node is scaled down.

          Default: 600

        • slurmCustomSettings (list) --

          Additional Slurm-specific configuration that directly maps to Slurm settings.

          • (dict) --

            Additional settings that directly map to Slurm settings.

            • parameterName (string) --

              Amazon Web Services PCS supports configuration of the following Slurm parameters:

            • parameterValue (string) --

              The values for the configured Slurm settings.

        • authKey (dict) --

          The shared Slurm key for authentication, also known as the cluster secret.

          • secretArn (string) --

            The Amazon Resource Name (ARN) of the the shared Slurm key.

          • secretVersion (string) --

            The version of the shared Slurm key.

        • accounting (dict) --

          The accounting configuration includes configurable settings for Slurm accounting.

          • mode (string) --

            The default value for mode is STANDARD. A value of STANDARD means Slurm accounting is enabled.

          • defaultPurgeTimeInDays (integer) --

            The default value for all purge settings for slurmdbd.conf. For more information, see the slurmdbd.conf documentation at SchedMD.

            The default value for defaultPurgeTimeInDays is -1.

            A value of -1 means there is no purge time and records persist as long as the cluster exists.

      • networking (dict) --

        The networking configuration for the cluster's control plane.

        • subnetIds (list) --

          The ID of the subnet where Amazon Web Services PCS creates an Elastic Network Interface (ENI) to enable communication between managed controllers and Amazon Web Services PCS resources. The subnet must have an available IP address, cannot reside in AWS Outposts, AWS Wavelength, or an AWS Local Zone.

          Example: subnet-abcd1234

          • (string) --

        • securityGroupIds (list) --

          The list of security group IDs associated with the Elastic Network Interface (ENI) created in subnets.

          The following rules are required:

          • Inbound rule 1

            • Protocol: All

            • Ports: All

            • Source: Self

          • Outbound rule 1

            • Protocol: All

            • Ports: All

            • Destination: 0.0.0.0/0 (IPv4) or ::/0 (IPv6)

          • Outbound rule 2

            • Protocol: All

            • Ports: All

            • Destination: Self

          • (string) --

        • networkType (string) --

          The IP address version the cluster uses. The default is IPV4.

      • endpoints (list) --

        The list of endpoints available for interaction with the scheduler.

        • (dict) --

          An endpoint available for interaction with the scheduler.

          • type (string) --

            Indicates the type of endpoint running at the specific IP address.

          • privateIpAddress (string) --

            For clusters that use IPv4, this is the endpoint's private IP address.

            Example: 10.1.2.3

            For clusters configured to use IPv6, this is an empty string.

          • publicIpAddress (string) --

            The endpoint's public IP address.

            Example: 192.0.2.1

          • ipv6Address (string) --

            The endpoint's IPv6 address.

            Example: 2001:db8::1

          • port (string) --

            The endpoint's connection port number.

            Example: 1234

      • errorInfo (list) --

        The list of errors that occurred during cluster provisioning.

        • (dict) --

          An error that occurred during resource creation.

          • code (string) --

            The short-form error code.

          • message (string) --

            The detailed error information.

GetCluster (updated) Link ¶
Changes (response)
{'cluster': {'endpoints': {'ipv6Address': 'string'},
             'networking': {'networkType': 'IPV4 | IPV6'}}}

Returns detailed information about a running cluster in your account. This API action provides networking information, endpoint information for communication with the scheduler, and provisioning status.

See also: AWS API Documentation

Request Syntax

client.get_cluster(
    clusterIdentifier='string'
)
type clusterIdentifier:

string

param clusterIdentifier:

[REQUIRED]

The name or ID of the cluster of the queue.

rtype:

dict

returns:

Response Syntax

{
    'cluster': {
        'name': 'string',
        'id': 'string',
        'arn': 'string',
        'status': 'CREATING'|'ACTIVE'|'UPDATING'|'DELETING'|'CREATE_FAILED'|'DELETE_FAILED'|'UPDATE_FAILED'|'SUSPENDING'|'SUSPENDED',
        'createdAt': datetime(2015, 1, 1),
        'modifiedAt': datetime(2015, 1, 1),
        'scheduler': {
            'type': 'SLURM',
            'version': 'string'
        },
        'size': 'SMALL'|'MEDIUM'|'LARGE',
        'slurmConfiguration': {
            'scaleDownIdleTimeInSeconds': 123,
            'slurmCustomSettings': [
                {
                    'parameterName': 'string',
                    'parameterValue': 'string'
                },
            ],
            'authKey': {
                'secretArn': 'string',
                'secretVersion': 'string'
            },
            'accounting': {
                'mode': 'STANDARD'|'NONE',
                'defaultPurgeTimeInDays': 123
            }
        },
        'networking': {
            'subnetIds': [
                'string',
            ],
            'securityGroupIds': [
                'string',
            ],
            'networkType': 'IPV4'|'IPV6'
        },
        'endpoints': [
            {
                'type': 'SLURMCTLD'|'SLURMDBD',
                'privateIpAddress': 'string',
                'publicIpAddress': 'string',
                'ipv6Address': 'string',
                'port': 'string'
            },
        ],
        'errorInfo': [
            {
                'code': 'string',
                'message': 'string'
            },
        ]
    }
}

Response Structure

  • (dict) --

    • cluster (dict) --

      The cluster resource.

      • name (string) --

        The name that identifies the cluster.

      • id (string) --

        The generated unique ID of the cluster.

      • arn (string) --

        The unique Amazon Resource Name (ARN) of the cluster.

      • status (string) --

        The provisioning status of the cluster.

      • createdAt (datetime) --

        The date and time the resource was created.

      • modifiedAt (datetime) --

        The date and time the resource was modified.

      • scheduler (dict) --

        The cluster management and job scheduling software associated with the cluster.

        • type (string) --

          The software Amazon Web Services PCS uses to manage cluster scaling and job scheduling.

        • version (string) --

          The version of the specified scheduling software that Amazon Web Services PCS uses to manage cluster scaling and job scheduling. For more information, see Slurm versions in Amazon Web Services PCS in the Amazon Web Services PCS User Guide.

          Valid Values: 23.11 | 24.05 | 24.11

      • size (string) --

        The size of the cluster.

        • SMALL: 32 compute nodes and 256 jobs

        • MEDIUM: 512 compute nodes and 8192 jobs

        • LARGE: 2048 compute nodes and 16,384 jobs

      • slurmConfiguration (dict) --

        Additional options related to the Slurm scheduler.

        • scaleDownIdleTimeInSeconds (integer) --

          The time (in seconds) before an idle node is scaled down.

          Default: 600

        • slurmCustomSettings (list) --

          Additional Slurm-specific configuration that directly maps to Slurm settings.

          • (dict) --

            Additional settings that directly map to Slurm settings.

            • parameterName (string) --

              Amazon Web Services PCS supports configuration of the following Slurm parameters:

            • parameterValue (string) --

              The values for the configured Slurm settings.

        • authKey (dict) --

          The shared Slurm key for authentication, also known as the cluster secret.

          • secretArn (string) --

            The Amazon Resource Name (ARN) of the the shared Slurm key.

          • secretVersion (string) --

            The version of the shared Slurm key.

        • accounting (dict) --

          The accounting configuration includes configurable settings for Slurm accounting.

          • mode (string) --

            The default value for mode is STANDARD. A value of STANDARD means Slurm accounting is enabled.

          • defaultPurgeTimeInDays (integer) --

            The default value for all purge settings for slurmdbd.conf. For more information, see the slurmdbd.conf documentation at SchedMD.

            The default value for defaultPurgeTimeInDays is -1.

            A value of -1 means there is no purge time and records persist as long as the cluster exists.

      • networking (dict) --

        The networking configuration for the cluster's control plane.

        • subnetIds (list) --

          The ID of the subnet where Amazon Web Services PCS creates an Elastic Network Interface (ENI) to enable communication between managed controllers and Amazon Web Services PCS resources. The subnet must have an available IP address, cannot reside in AWS Outposts, AWS Wavelength, or an AWS Local Zone.

          Example: subnet-abcd1234

          • (string) --

        • securityGroupIds (list) --

          The list of security group IDs associated with the Elastic Network Interface (ENI) created in subnets.

          The following rules are required:

          • Inbound rule 1

            • Protocol: All

            • Ports: All

            • Source: Self

          • Outbound rule 1

            • Protocol: All

            • Ports: All

            • Destination: 0.0.0.0/0 (IPv4) or ::/0 (IPv6)

          • Outbound rule 2

            • Protocol: All

            • Ports: All

            • Destination: Self

          • (string) --

        • networkType (string) --

          The IP address version the cluster uses. The default is IPV4.

      • endpoints (list) --

        The list of endpoints available for interaction with the scheduler.

        • (dict) --

          An endpoint available for interaction with the scheduler.

          • type (string) --

            Indicates the type of endpoint running at the specific IP address.

          • privateIpAddress (string) --

            For clusters that use IPv4, this is the endpoint's private IP address.

            Example: 10.1.2.3

            For clusters configured to use IPv6, this is an empty string.

          • publicIpAddress (string) --

            The endpoint's public IP address.

            Example: 192.0.2.1

          • ipv6Address (string) --

            The endpoint's IPv6 address.

            Example: 2001:db8::1

          • port (string) --

            The endpoint's connection port number.

            Example: 1234

      • errorInfo (list) --

        The list of errors that occurred during cluster provisioning.

        • (dict) --

          An error that occurred during resource creation.

          • code (string) --

            The short-form error code.

          • message (string) --

            The detailed error information.

RegisterComputeNodeGroupInstance (updated) Link ¶
Changes (response)
{'endpoints': {'ipv6Address': 'string'}}

Amazon Web Services PCS uses this API action to register the compute nodes it launches in your account.

See also: AWS API Documentation

Request Syntax

client.register_compute_node_group_instance(
    clusterIdentifier='string',
    bootstrapId='string'
)
type clusterIdentifier:

string

param clusterIdentifier:

[REQUIRED]

The name or ID of the cluster to register the compute node group instance in.

type bootstrapId:

string

param bootstrapId:

[REQUIRED]

The client-generated token to allow for retries.

rtype:

dict

returns:

Response Syntax

{
    'nodeID': 'string',
    'sharedSecret': 'string',
    'endpoints': [
        {
            'type': 'SLURMCTLD'|'SLURMDBD',
            'privateIpAddress': 'string',
            'publicIpAddress': 'string',
            'ipv6Address': 'string',
            'port': 'string'
        },
    ]
}

Response Structure

  • (dict) --

    • nodeID (string) --

      The scheduler node ID for this instance.

    • sharedSecret (string) --

      For the Slurm scheduler, this is the shared Munge key the scheduler uses to authenticate compute node group instances.

    • endpoints (list) --

      The list of endpoints available for interaction with the scheduler.

      • (dict) --

        An endpoint available for interaction with the scheduler.

        • type (string) --

          Indicates the type of endpoint running at the specific IP address.

        • privateIpAddress (string) --

          For clusters that use IPv4, this is the endpoint's private IP address.

          Example: 10.1.2.3

          For clusters configured to use IPv6, this is an empty string.

        • publicIpAddress (string) --

          The endpoint's public IP address.

          Example: 192.0.2.1

        • ipv6Address (string) --

          The endpoint's IPv6 address.

          Example: 2001:db8::1

        • port (string) --

          The endpoint's connection port number.

          Example: 1234