DedicatedResources

A description of resources that are dedicated to a DeployedModel, and that need a higher degree of manual configuration.

Fields
machineSpec object (MachineSpec)

Required. Immutable. The specification of a single machine used by the prediction.

minReplicaCount integer

Required. Immutable. The minimum number of machine replicas this DeployedModel will be always deployed on. This value must be greater than or equal to 1.

If traffic against the DeployedModel increases, it may dynamically be deployed onto more replicas, and as traffic decreases, some of these extra replicas may be freed.

maxReplicaCount integer

Immutable. The maximum number of replicas this DeployedModel may be deployed on when the traffic against it increases. If the requested value is too large, the deployment will error, but if deployment succeeds then the ability to scale the model to that many replicas is guaranteed (barring service outages). If traffic against the DeployedModel increases beyond what its replicas at maximum may handle, a portion of the traffic will be dropped. If this value is not provided, will use minReplicaCount as the default value.

The value of this field impacts the charge against Vertex CPU and GPU quotas. Specifically, you will be charged for (maxReplicaCount * number of cores in the selected machine type) and (maxReplicaCount * number of GPUs per replica in the selected machine type).

autoscalingMetricSpecs[] object (AutoscalingMetricSpec)

Immutable. The metric specifications that overrides a resource utilization metric (CPU utilization, accelerator's duty cycle, and so on) target value (default to 60 if not set). At most one entry is allowed per metric.

If machineSpec.accelerator_count is above 0, the autoscaling will be based on both CPU utilization and accelerator's duty cycle metrics and scale up when either metrics exceeds its target value while scale down if both metrics are under their target value. The default target value is 60 for both metrics.

If machineSpec.accelerator_count is 0, the autoscaling will be based on CPU utilization metric only with default target value 60 if not explicitly set.

For example, in the case of Online Prediction, if you want to override target CPU utilization to 80, you should set autoscalingMetricSpecs.metric_name to aiplatform.googleapis.com/prediction/online/cpu/utilization and autoscalingMetricSpecs.target to 80.

spot boolean

Optional. If true, schedule the deployment workload on spot VMs.

JSON representation
{
  "machineSpec": {
    object (MachineSpec)
  },
  "minReplicaCount": integer,
  "maxReplicaCount": integer,
  "autoscalingMetricSpecs": [
    {
      object (AutoscalingMetricSpec)
    }
  ],
  "spot": boolean
}

AutoscalingMetricSpec

The metric specification that defines the target resource utilization (CPU utilization, accelerator's duty cycle, and so on) for calculating the desired replica count.

Fields
metricName string

Required. The resource metric name. Supported metrics:

  • For Online Prediction:
  • aiplatform.googleapis.com/prediction/online/accelerator/duty_cycle
  • aiplatform.googleapis.com/prediction/online/cpu/utilization
target integer

The target resource utilization in percentage (1% - 100%) for the given metric; once the real usage deviates from the target by a certain percentage, the machine replicas change. The default value is 60 (representing 60%) if not provided.

JSON representation
{
  "metricName": string,
  "target": integer
}