interface ClusterInstanceGroupProperty
| Language | Type name |
|---|---|
 .NET | Amazon.CDK.Mixins.Preview.AWS.SageMaker.Mixins.CfnClusterPropsMixin.ClusterInstanceGroupProperty |
 Go | github.com/aws/aws-cdk-go/awscdkmixinspreview/v2/awssagemaker/mixins#CfnClusterPropsMixin_ClusterInstanceGroupProperty |
 Java | software.amazon.awscdk.mixins.preview.services.sagemaker.mixins.CfnClusterPropsMixin.ClusterInstanceGroupProperty |
 Python | aws_cdk.mixins_preview.aws_sagemaker.mixins.CfnClusterPropsMixin.ClusterInstanceGroupProperty |
 TypeScript | @aws-cdk/mixins-preview » aws_sagemaker » mixins » CfnClusterPropsMixin » ClusterInstanceGroupProperty |
The configuration information of the instance group within the HyperPod cluster.
Example
// The code below shows an example of how to instantiate this type.
// The values are placeholders you should change.
import { mixins as sagemaker_mixins } from '@aws-cdk/mixins-preview/aws-sagemaker';
declare const onDemand: any;
declare const spot: any;
const clusterInstanceGroupProperty: sagemaker_mixins.CfnClusterPropsMixin.ClusterInstanceGroupProperty = {
capacityRequirements: {
onDemand: onDemand,
spot: spot,
},
currentCount: 123,
executionRole: 'executionRole',
imageId: 'imageId',
instanceCount: 123,
instanceGroupName: 'instanceGroupName',
instanceStorageConfigs: [{
ebsVolumeConfig: {
rootVolume: false,
volumeKmsKeyId: 'volumeKmsKeyId',
volumeSizeInGb: 123,
},
}],
instanceType: 'instanceType',
kubernetesConfig: {
labels: {
labelsKey: 'labels',
},
taints: [{
effect: 'effect',
key: 'key',
value: 'value',
}],
},
lifeCycleConfig: {
onCreate: 'onCreate',
sourceS3Uri: 'sourceS3Uri',
},
minInstanceCount: 123,
onStartDeepHealthChecks: ['onStartDeepHealthChecks'],
overrideVpcConfig: {
securityGroupIds: ['securityGroupIds'],
subnets: ['subnets'],
},
scheduledUpdateConfig: {
deploymentConfig: {
autoRollbackConfiguration: [{
alarmName: 'alarmName',
}],
rollingUpdatePolicy: {
maximumBatchSize: {
type: 'type',
value: 123,
},
rollbackMaximumBatchSize: {
type: 'type',
value: 123,
},
},
waitIntervalInSeconds: 123,
},
scheduleExpression: 'scheduleExpression',
},
threadsPerCore: 123,
trainingPlanArn: 'trainingPlanArn',
};
Properties
| Name | Type | Description |
|---|---|---|
| capacity | IResolvable | Cluster | Specifies the capacity requirements configuration for an instance group. |
| current | number | The number of instances that are currently in the instance group of a SageMaker HyperPod cluster. |
| execution | string | The execution role for the instance group to assume. |
| image | string | AMI Id to be used for launching EC2 instances - HyperPodPublicAmiId or CustomAmiId. |
| instance | number | The number of instances in an instance group of the SageMaker HyperPod cluster. |
| instance | string | The name of the instance group of a SageMaker HyperPod cluster. |
| instance | IResolvable | (IResolvable | Cluster)[] | The configurations of additional storage specified to the instance group where the instance (node) is launched. |
| instance | string | The instance type of the instance group of a SageMaker HyperPod cluster. |
| kubernetes | IResolvable | Cluster | Kubernetes configuration for cluster nodes including labels and taints. |
| life | IResolvable | Cluster | The lifecycle configuration for a SageMaker HyperPod cluster. |
| min | number | The minimum number of instances required for the instance group to be InService. |
| on | string[] | A flag indicating whether deep health checks should be performed when the HyperPod cluster instance group is created or updated. |
| override | IResolvable | Vpc | The customized Amazon VPC configuration at the instance group level that overrides the default Amazon VPC configuration of the SageMaker HyperPod cluster. |
| scheduled | IResolvable | Scheduled | The configuration object of the schedule that SageMaker follows when updating the AMI. |
| threads | number | The number of threads per CPU core you specified under CreateCluster . |
| training | string | The Amazon Resource Name (ARN) of the training plan to use for this cluster instance group. |
capacityRequirements?
Type:
IResolvable | Cluster
(optional)
Specifies the capacity requirements configuration for an instance group.
currentCount?
Type:
number
(optional)
The number of instances that are currently in the instance group of a SageMaker HyperPod cluster.
executionRole?
Type:
string
(optional)
The execution role for the instance group to assume.
imageId?
Type:
string
(optional)
AMI Id to be used for launching EC2 instances - HyperPodPublicAmiId or CustomAmiId.
instanceCount?
Type:
number
(optional)
The number of instances in an instance group of the SageMaker HyperPod cluster.
instanceGroupName?
Type:
string
(optional)
The name of the instance group of a SageMaker HyperPod cluster.
instanceStorageConfigs?
Type:
IResolvable | (IResolvable | Cluster)[]
(optional)
The configurations of additional storage specified to the instance group where the instance (node) is launched.
instanceType?
Type:
string
(optional)
The instance type of the instance group of a SageMaker HyperPod cluster.
kubernetesConfig?
Type:
IResolvable | Cluster
(optional)
Kubernetes configuration for cluster nodes including labels and taints.
lifeCycleConfig?
Type:
IResolvable | Cluster
(optional)
The lifecycle configuration for a SageMaker HyperPod cluster.
minInstanceCount?
Type:
number
(optional)
The minimum number of instances required for the instance group to be InService.
MinInstanceCount must be less than or equal to InstanceCount.
onStartDeepHealthChecks?
Type:
string[]
(optional)
A flag indicating whether deep health checks should be performed when the HyperPod cluster instance group is created or updated.
Deep health checks are comprehensive, invasive tests that validate the health of the underlying hardware and infrastructure components.
overrideVpcConfig?
Type:
IResolvable | Vpc
(optional)
The customized Amazon VPC configuration at the instance group level that overrides the default Amazon VPC configuration of the SageMaker HyperPod cluster.
scheduledUpdateConfig?
Type:
IResolvable | Scheduled
(optional)
The configuration object of the schedule that SageMaker follows when updating the AMI.
threadsPerCore?
Type:
number
(optional)
The number of threads per CPU core you specified under CreateCluster .
trainingPlanArn?
Type:
string
(optional)
The Amazon Resource Name (ARN) of the training plan to use for this cluster instance group.
For more information about how to reserve GPU capacity for your SageMaker HyperPod clusters using Amazon SageMaker Training Plan, see CreateTrainingPlan.