AiDeploymentResourcePool class

'DeploymentResourcePool can be shared by multiple deployed models, whose underlying specification consists of dedicated resources.'

To get more information about DeploymentResourcePool, see:

API documentation

Example Usage

Vertex Ai Deployment Resource Pool

import * as pulumi from "@pulumi/pulumi";
import * as gcp from "@pulumi/gcp";

const deploymentResourcePool = new gcp.vertex.AiDeploymentResourcePool("deployment_resource_pool", {
    region: "us-central1",
    name: "example-deployment-resource-pool",
    dedicatedResources: {
        machineSpec: {
            machineType: "n1-standard-4",
            acceleratorType: "NVIDIA_TESLA_P4",
            acceleratorCount: 1,
        },
        minReplicaCount: 1,
        maxReplicaCount: 2,
        autoscalingMetricSpecs: [{
            metricName: "aiplatform.googleapis.com/prediction/online/accelerator/duty_cycle",
            target: 60,
        }],
    },
});

import pulumi
import pulumi_gcp as gcp

deployment_resource_pool = gcp.vertex.AiDeploymentResourcePool("deployment_resource_pool",
    region="us-central1",
    name="example-deployment-resource-pool",
    dedicated_resources={
        "machine_spec": {
            "machine_type": "n1-standard-4",
            "accelerator_type": "NVIDIA_TESLA_P4",
            "accelerator_count": 1,
        },
        "min_replica_count": 1,
        "max_replica_count": 2,
        "autoscaling_metric_specs": [{
            "metric_name": "aiplatform.googleapis.com/prediction/online/accelerator/duty_cycle",
            "target": 60,
        }],
    })

using System.Collections.Generic;
using System.Linq;
using Pulumi;
using Gcp = Pulumi.Gcp;

return await Deployment.RunAsync(() =>
{
    var deploymentResourcePool = new Gcp.Vertex.AiDeploymentResourcePool("deployment_resource_pool", new()
    {
        Region = "us-central1",
        Name = "example-deployment-resource-pool",
        DedicatedResources = new Gcp.Vertex.Inputs.AiDeploymentResourcePoolDedicatedResourcesArgs
        {
            MachineSpec = new Gcp.Vertex.Inputs.AiDeploymentResourcePoolDedicatedResourcesMachineSpecArgs
            {
                MachineType = "n1-standard-4",
                AcceleratorType = "NVIDIA_TESLA_P4",
                AcceleratorCount = 1,
            },
            MinReplicaCount = 1,
            MaxReplicaCount = 2,
            AutoscalingMetricSpecs = new[]
            {
                new Gcp.Vertex.Inputs.AiDeploymentResourcePoolDedicatedResourcesAutoscalingMetricSpecArgs
                {
                    MetricName = "aiplatform.googleapis.com/prediction/online/accelerator/duty_cycle",
                    Target = 60,
                },
            },
        },
    });

});

package main

import (
	"github.com/pulumi/pulumi-gcp/sdk/v9/go/gcp/vertex"
	"github.com/pulumi/pulumi/sdk/v3/go/pulumi"
)

func main() {
	pulumi.Run(func(ctx *pulumi.Context) error {
		_, err := vertex.NewAiDeploymentResourcePool(ctx, "deployment_resource_pool", &vertex.AiDeploymentResourcePoolArgs{
			Region: pulumi.String("us-central1"),
			Name:   pulumi.String("example-deployment-resource-pool"),
			DedicatedResources: &vertex.AiDeploymentResourcePoolDedicatedResourcesArgs{
				MachineSpec: &vertex.AiDeploymentResourcePoolDedicatedResourcesMachineSpecArgs{
					MachineType:      pulumi.String("n1-standard-4"),
					AcceleratorType:  pulumi.String("NVIDIA_TESLA_P4"),
					AcceleratorCount: pulumi.Int(1),
				},
				MinReplicaCount: pulumi.Int(1),
				MaxReplicaCount: pulumi.Int(2),
				AutoscalingMetricSpecs: vertex.AiDeploymentResourcePoolDedicatedResourcesAutoscalingMetricSpecArray{
					&vertex.AiDeploymentResourcePoolDedicatedResourcesAutoscalingMetricSpecArgs{
						MetricName: pulumi.String("aiplatform.googleapis.com/prediction/online/accelerator/duty_cycle"),
						Target:     pulumi.Int(60),
					},
				},
			},
		})
		if err != nil {
			return err
		}
		return nil
	})
}

pulumi {
  required_providers {
    gcp = {
      source = "pulumi/gcp"
    }
  }
}

resource "gcp_vertex_aideploymentresourcepool" "deployment_resource_pool" {
  region = "us-central1"
  name   = "example-deployment-resource-pool"
  dedicated_resources = {
    machine_spec = {
      machine_type      = "n1-standard-4"
      accelerator_type  = "NVIDIA_TESLA_P4"
      accelerator_count = 1
    }
    min_replica_count = 1
    max_replica_count = 2
    autoscaling_metric_specs = [{
      "metricName" = "aiplatform.googleapis.com/prediction/online/accelerator/duty_cycle"
      "target"     = 60
    }]
  }
}

package generated_program;

import com.pulumi.Context;
import com.pulumi.Pulumi;
import com.pulumi.core.Output;
import com.pulumi.gcp.vertex.AiDeploymentResourcePool;
import com.pulumi.gcp.vertex.AiDeploymentResourcePoolArgs;
import com.pulumi.gcp.vertex.inputs.AiDeploymentResourcePoolDedicatedResourcesArgs;
import com.pulumi.gcp.vertex.inputs.AiDeploymentResourcePoolDedicatedResourcesMachineSpecArgs;
import com.pulumi.gcp.vertex.inputs.AiDeploymentResourcePoolDedicatedResourcesAutoscalingMetricSpecArgs;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Map;
import java.io.File;
import java.nio.file.Files;
import java.nio.file.Paths;

public class App {
    public static void main(String[] args) {
        Pulumi.run(App::stack);
    }

    public static void stack(Context ctx) {
        var deploymentResourcePool = new AiDeploymentResourcePool("deploymentResourcePool", AiDeploymentResourcePoolArgs.builder()
            .region("us-central1")
            .name("example-deployment-resource-pool")
            .dedicatedResources(AiDeploymentResourcePoolDedicatedResourcesArgs.builder()
                .machineSpec(AiDeploymentResourcePoolDedicatedResourcesMachineSpecArgs.builder()
                    .machineType("n1-standard-4")
                    .acceleratorType("NVIDIA_TESLA_P4")
                    .acceleratorCount(1)
                    .build())
                .minReplicaCount(1)
                .maxReplicaCount(2)
                .autoscalingMetricSpecs(AiDeploymentResourcePoolDedicatedResourcesAutoscalingMetricSpecArgs.builder()
                    .metricName("aiplatform.googleapis.com/prediction/online/accelerator/duty_cycle")
                    .target(60)
                    .build())
                .build())
            .build());

    }
}

resources:
  deploymentResourcePool:
    type: gcp:vertex:AiDeploymentResourcePool
    name: deployment_resource_pool
    properties:
      region: us-central1
      name: example-deployment-resource-pool
      dedicatedResources:
        machineSpec:
          machineType: n1-standard-4
          acceleratorType: NVIDIA_TESLA_P4
          acceleratorCount: 1
        minReplicaCount: 1
        maxReplicaCount: 2
        autoscalingMetricSpecs:
          - metricName: aiplatform.googleapis.com/prediction/online/accelerator/duty_cycle
            target: 60

Import

DeploymentResourcePool can be imported using any of these accepted formats:

projects/{{project}}/locations/{{region}}/deploymentResourcePools/{{name}}
{{project}}/{{region}}/{{name}}
{{region}}/{{name}}
{{name}}

When using the pulumi import command, DeploymentResourcePool can be imported using one of the formats above. For example:

$ pulumi import gcp:vertex/aiDeploymentResourcePool:AiDeploymentResourcePool default projects/{{project}}/locations/{{region}}/deploymentResourcePools/{{name}}
$ pulumi import gcp:vertex/aiDeploymentResourcePool:AiDeploymentResourcePool default {{project}}/{{region}}/{{name}}
$ pulumi import gcp:vertex/aiDeploymentResourcePool:AiDeploymentResourcePool default {{region}}/{{name}}
$ pulumi import gcp:vertex/aiDeploymentResourcePool:AiDeploymentResourcePool default {{name}}

Constructors

AiDeploymentResourcePool(String name, {AiDeploymentResourcePoolArgs? args, CustomResourceOptions? options}): Creates a new AiDeploymentResourcePool. name The Pulumi resource name. args Arguments used to configure this AiDeploymentResourcePool. The set of arguments for AiDeploymentResourcePool. options Resource options controlling this resource's behavior.
AiDeploymentResourcePool.reference(String urn): Creates a typed reference to an existing AiDeploymentResourcePool resource.

Properties

childResources → Set<Resource>: finalinherited
completionSources ↔ Map<String, IOutputCompletionSource>: latefinalinherited
createTime ↔ Output<String>: A timestamp in RFC3339 UTC "Zulu" format, with nanosecond resolution and up to nine fractional digits.
latefinal
dedicatedResources ↔ Output<AiDeploymentResourcePoolDedicatedResources?>: The underlying dedicated resources that the deployment resource pool uses. Structure is documented below.
latefinal
deletionPolicy ↔ Output<String>: Whether Terraform will be prevented from destroying the resource. Defaults to DELETE. When a 'terraform destroy' or 'pulumi up' would delete the resource, the command will fail if this field is set to "PREVENT" in Terraform state. When set to "ABANDON", the command will remove the resource from Terraform management without updating or deleting the resource in the API. When set to "DELETE", deleting the resource is allowed.
latefinal
hashCode → int: The hash code for this object.
no setterinherited
id ↔ Output<String>: getter/setter pairinherited
isCustom → bool: Returns whether this resource is provider-managed.
no setterinherited
isProtected → bool: Returns whether this resource is protected from deletion.
no setterinherited
isRemote → bool: Whether this resource is registered as remote.
no setterinherited
isResourceReference → bool: Whether this instance represents a resource value returned over RPC.
finalinherited
name ↔ Output<String>: The resource name of deployment resource pool. The maximum length is 63 characters, and valid characters are /^a-z?$/.
latefinal
project ↔ Output<String>: The ID of the project in which the resource belongs. If it is not provided, the provider project is used.
latefinal
region ↔ Output<String?>: The region of deployment resource pool. eg us-central1
latefinal
resourceTransforms → List<ResourceTransform>: Inherited/explicit async transforms.
no setterinherited
runtimeType → Type: A representation of the runtime type of the object.
no setterinherited
transformations → List<ResourceTransformation>: Inherited/explicit legacy transformations.
no setterinherited
urn ↔ Output<String>: latefinalinherited

Methods

failId(Object error) → void: Completes this resource ID with an error when registration fails.
inherited
failOutputs(Object error) → void: Completes all output properties with error.
inherited
failUrn(Object error) → void: Completes this resource URN with an error when registration fails.
inherited
getProvider(String moduleMember) → ProviderResource?: Returns provider for moduleMember's package, if configured.
inherited
getResourceName() → String: Returns this resource's logical name.
inherited
getResourceType() → String: Returns this resource's Pulumi type token.
inherited
noSuchMethod(Invocation invocation) → dynamic: Invoked when a nonexistent method or property is accessed.
inherited
registerOutput<T>(String propertyName, {Object? decoder(Object?)?, bool isSecret = false}) → Output<T>: Registers a dynamic output property for this resource.
inherited
resolveId(String? value, {required bool isKnown}) → void: Resolves the provider-assigned ID for this resource.
inherited
resolveOutputs(Struct outputs) → void: Resolves all output properties from a monitor response payload.
inherited
resolveUrn(String value) → void: Resolves this resource's URN once assigned by the engine.
inherited
serializeProperties(Map<String, dynamic> properties) → Future<Struct>: Serializes resource properties for RPC transmission.
inherited
toString() → String: A string representation of this object.
inherited

Operators

operator ==(Object other) → bool: The equality operator.
inherited

Static Methods

get(String name, Input<String> id, {AiDeploymentResourcePoolState? state, CustomResourceOptions? options}) → AiDeploymentResourcePool: Gets an existing AiDeploymentResourcePool resource's state with the given name and id.