Manage GCP Data Fusion Instance IAM Bindings

The gcp:datafusion/instance:Instance resource, part of the Pulumi GCP provider, provisions a Cloud Data Fusion instance: the managed data integration platform with its tier, region, and optional networking and encryption configuration. This guide focuses on four capabilities: basic instance creation, private networking with VPC peering, customer-managed encryption, and event publishing to Pub/Sub.

Data Fusion instances may reference VPC networks, KMS keys, service accounts, and Pub/Sub topics that must exist separately. The examples are intentionally small. Combine them with your own networking, encryption, and monitoring infrastructure.

Create a basic Data Fusion instance

Most deployments start with a minimal instance that provides the core data integration platform.

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

const basicInstance = new gcp.datafusion.Instance("basic_instance", {
    name: "my-instance",
    region: "us-central1",
    type: "BASIC",
});
import pulumi
import pulumi_gcp as gcp

basic_instance = gcp.datafusion.Instance("basic_instance",
    name="my-instance",
    region="us-central1",
    type="BASIC")
package main

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

func main() {
	pulumi.Run(func(ctx *pulumi.Context) error {
		_, err := datafusion.NewInstance(ctx, "basic_instance", &datafusion.InstanceArgs{
			Name:   pulumi.String("my-instance"),
			Region: pulumi.String("us-central1"),
			Type:   pulumi.String("BASIC"),
		})
		if err != nil {
			return err
		}
		return nil
	})
}
using System.Collections.Generic;
using System.Linq;
using Pulumi;
using Gcp = Pulumi.Gcp;

return await Deployment.RunAsync(() => 
{
    var basicInstance = new Gcp.DataFusion.Instance("basic_instance", new()
    {
        Name = "my-instance",
        Region = "us-central1",
        Type = "BASIC",
    });

});
package generated_program;

import com.pulumi.Context;
import com.pulumi.Pulumi;
import com.pulumi.core.Output;
import com.pulumi.gcp.datafusion.Instance;
import com.pulumi.gcp.datafusion.InstanceArgs;
import java.util.List;
import java.util.ArrayList;
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 basicInstance = new Instance("basicInstance", InstanceArgs.builder()
            .name("my-instance")
            .region("us-central1")
            .type("BASIC")
            .build());

    }
}
resources:
  basicInstance:
    type: gcp:datafusion:Instance
    name: basic_instance
    properties:
      name: my-instance
      region: us-central1
      type: BASIC

The type property sets the instance tier (BASIC, ENTERPRISE, or DEVELOPER), which determines available features and capacity. The name and region properties identify where the instance runs. Without additional configuration, the instance is publicly accessible and uses Google-managed encryption.

Configure monitoring and private networking

Production deployments often isolate Data Fusion from the public internet and enable observability through Stackdriver.

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

const _default = gcp.appengine.getDefaultServiceAccount({});
const network = new gcp.compute.Network("network", {name: "datafusion-full-network"});
const privateIpAlloc = new gcp.compute.GlobalAddress("private_ip_alloc", {
    name: "datafusion-ip-alloc",
    addressType: "INTERNAL",
    purpose: "VPC_PEERING",
    prefixLength: 22,
    network: network.id,
});
const extendedInstance = new gcp.datafusion.Instance("extended_instance", {
    name: "my-instance",
    description: "My Data Fusion instance",
    displayName: "My Data Fusion instance",
    region: "us-central1",
    type: "BASIC",
    enableStackdriverLogging: true,
    enableStackdriverMonitoring: true,
    privateInstance: true,
    dataprocServiceAccount: _default.then(_default => _default.email),
    labels: {
        example_key: "example_value",
    },
    networkConfig: {
        network: "default",
        ipAllocation: pulumi.interpolate`${privateIpAlloc.address}/${privateIpAlloc.prefixLength}`,
    },
    accelerators: [{
        acceleratorType: "CDC",
        state: "ENABLED",
    }],
});
import pulumi
import pulumi_gcp as gcp

default = gcp.appengine.get_default_service_account()
network = gcp.compute.Network("network", name="datafusion-full-network")
private_ip_alloc = gcp.compute.GlobalAddress("private_ip_alloc",
    name="datafusion-ip-alloc",
    address_type="INTERNAL",
    purpose="VPC_PEERING",
    prefix_length=22,
    network=network.id)
extended_instance = gcp.datafusion.Instance("extended_instance",
    name="my-instance",
    description="My Data Fusion instance",
    display_name="My Data Fusion instance",
    region="us-central1",
    type="BASIC",
    enable_stackdriver_logging=True,
    enable_stackdriver_monitoring=True,
    private_instance=True,
    dataproc_service_account=default.email,
    labels={
        "example_key": "example_value",
    },
    network_config={
        "network": "default",
        "ip_allocation": pulumi.Output.all(
            address=private_ip_alloc.address,
            prefix_length=private_ip_alloc.prefix_length
).apply(lambda resolved_outputs: f"{resolved_outputs['address']}/{resolved_outputs['prefix_length']}")
,
    },
    accelerators=[{
        "accelerator_type": "CDC",
        "state": "ENABLED",
    }])
package main

import (
	"fmt"

	"github.com/pulumi/pulumi-gcp/sdk/v9/go/gcp/appengine"
	"github.com/pulumi/pulumi-gcp/sdk/v9/go/gcp/compute"
	"github.com/pulumi/pulumi-gcp/sdk/v9/go/gcp/datafusion"
	"github.com/pulumi/pulumi/sdk/v3/go/pulumi"
)

func main() {
	pulumi.Run(func(ctx *pulumi.Context) error {
		_default, err := appengine.GetDefaultServiceAccount(ctx, &appengine.GetDefaultServiceAccountArgs{}, nil)
		if err != nil {
			return err
		}
		network, err := compute.NewNetwork(ctx, "network", &compute.NetworkArgs{
			Name: pulumi.String("datafusion-full-network"),
		})
		if err != nil {
			return err
		}
		privateIpAlloc, err := compute.NewGlobalAddress(ctx, "private_ip_alloc", &compute.GlobalAddressArgs{
			Name:         pulumi.String("datafusion-ip-alloc"),
			AddressType:  pulumi.String("INTERNAL"),
			Purpose:      pulumi.String("VPC_PEERING"),
			PrefixLength: pulumi.Int(22),
			Network:      network.ID(),
		})
		if err != nil {
			return err
		}
		_, err = datafusion.NewInstance(ctx, "extended_instance", &datafusion.InstanceArgs{
			Name:                        pulumi.String("my-instance"),
			Description:                 pulumi.String("My Data Fusion instance"),
			DisplayName:                 pulumi.String("My Data Fusion instance"),
			Region:                      pulumi.String("us-central1"),
			Type:                        pulumi.String("BASIC"),
			EnableStackdriverLogging:    pulumi.Bool(true),
			EnableStackdriverMonitoring: pulumi.Bool(true),
			PrivateInstance:             pulumi.Bool(true),
			DataprocServiceAccount:      pulumi.String(_default.Email),
			Labels: pulumi.StringMap{
				"example_key": pulumi.String("example_value"),
			},
			NetworkConfig: &datafusion.InstanceNetworkConfigArgs{
				Network: pulumi.String("default"),
				IpAllocation: pulumi.All(privateIpAlloc.Address, privateIpAlloc.PrefixLength).ApplyT(func(_args []interface{}) (string, error) {
					address := _args[0].(string)
					prefixLength := _args[1].(int)
					return fmt.Sprintf("%v/%v", address, prefixLength), nil
				}).(pulumi.StringOutput),
			},
			Accelerators: datafusion.InstanceAcceleratorArray{
				&datafusion.InstanceAcceleratorArgs{
					AcceleratorType: pulumi.String("CDC"),
					State:           pulumi.String("ENABLED"),
				},
			},
		})
		if err != nil {
			return err
		}
		return nil
	})
}
using System.Collections.Generic;
using System.Linq;
using Pulumi;
using Gcp = Pulumi.Gcp;

return await Deployment.RunAsync(() => 
{
    var @default = Gcp.AppEngine.GetDefaultServiceAccount.Invoke();

    var network = new Gcp.Compute.Network("network", new()
    {
        Name = "datafusion-full-network",
    });

    var privateIpAlloc = new Gcp.Compute.GlobalAddress("private_ip_alloc", new()
    {
        Name = "datafusion-ip-alloc",
        AddressType = "INTERNAL",
        Purpose = "VPC_PEERING",
        PrefixLength = 22,
        Network = network.Id,
    });

    var extendedInstance = new Gcp.DataFusion.Instance("extended_instance", new()
    {
        Name = "my-instance",
        Description = "My Data Fusion instance",
        DisplayName = "My Data Fusion instance",
        Region = "us-central1",
        Type = "BASIC",
        EnableStackdriverLogging = true,
        EnableStackdriverMonitoring = true,
        PrivateInstance = true,
        DataprocServiceAccount = @default.Apply(@default => @default.Apply(getDefaultServiceAccountResult => getDefaultServiceAccountResult.Email)),
        Labels = 
        {
            { "example_key", "example_value" },
        },
        NetworkConfig = new Gcp.DataFusion.Inputs.InstanceNetworkConfigArgs
        {
            Network = "default",
            IpAllocation = Output.Tuple(privateIpAlloc.Address, privateIpAlloc.PrefixLength).Apply(values =>
            {
                var address = values.Item1;
                var prefixLength = values.Item2;
                return $"{address}/{prefixLength}";
            }),
        },
        Accelerators = new[]
        {
            new Gcp.DataFusion.Inputs.InstanceAcceleratorArgs
            {
                AcceleratorType = "CDC",
                State = "ENABLED",
            },
        },
    });

});
package generated_program;

import com.pulumi.Context;
import com.pulumi.Pulumi;
import com.pulumi.core.Output;
import com.pulumi.gcp.appengine.AppengineFunctions;
import com.pulumi.gcp.appengine.inputs.GetDefaultServiceAccountArgs;
import com.pulumi.gcp.compute.Network;
import com.pulumi.gcp.compute.NetworkArgs;
import com.pulumi.gcp.compute.GlobalAddress;
import com.pulumi.gcp.compute.GlobalAddressArgs;
import com.pulumi.gcp.datafusion.Instance;
import com.pulumi.gcp.datafusion.InstanceArgs;
import com.pulumi.gcp.datafusion.inputs.InstanceNetworkConfigArgs;
import com.pulumi.gcp.datafusion.inputs.InstanceAcceleratorArgs;
import java.util.List;
import java.util.ArrayList;
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) {
        final var default = AppengineFunctions.getDefaultServiceAccount(GetDefaultServiceAccountArgs.builder()
            .build());

        var network = new Network("network", NetworkArgs.builder()
            .name("datafusion-full-network")
            .build());

        var privateIpAlloc = new GlobalAddress("privateIpAlloc", GlobalAddressArgs.builder()
            .name("datafusion-ip-alloc")
            .addressType("INTERNAL")
            .purpose("VPC_PEERING")
            .prefixLength(22)
            .network(network.id())
            .build());

        var extendedInstance = new Instance("extendedInstance", InstanceArgs.builder()
            .name("my-instance")
            .description("My Data Fusion instance")
            .displayName("My Data Fusion instance")
            .region("us-central1")
            .type("BASIC")
            .enableStackdriverLogging(true)
            .enableStackdriverMonitoring(true)
            .privateInstance(true)
            .dataprocServiceAccount(default_.email())
            .labels(Map.of("example_key", "example_value"))
            .networkConfig(InstanceNetworkConfigArgs.builder()
                .network("default")
                .ipAllocation(Output.tuple(privateIpAlloc.address(), privateIpAlloc.prefixLength()).applyValue(values -> {
                    var address = values.t1;
                    var prefixLength = values.t2;
                    return String.format("%s/%s", address,prefixLength);
                }))
                .build())
            .accelerators(InstanceAcceleratorArgs.builder()
                .acceleratorType("CDC")
                .state("ENABLED")
                .build())
            .build());

    }
}
resources:
  extendedInstance:
    type: gcp:datafusion:Instance
    name: extended_instance
    properties:
      name: my-instance
      description: My Data Fusion instance
      displayName: My Data Fusion instance
      region: us-central1
      type: BASIC
      enableStackdriverLogging: true
      enableStackdriverMonitoring: true
      privateInstance: true
      dataprocServiceAccount: ${default.email}
      labels:
        example_key: example_value
      networkConfig:
        network: default
        ipAllocation: ${privateIpAlloc.address}/${privateIpAlloc.prefixLength}
      accelerators:
        - acceleratorType: CDC
          state: ENABLED
  network:
    type: gcp:compute:Network
    properties:
      name: datafusion-full-network
  privateIpAlloc:
    type: gcp:compute:GlobalAddress
    name: private_ip_alloc
    properties:
      name: datafusion-ip-alloc
      addressType: INTERNAL
      purpose: VPC_PEERING
      prefixLength: 22
      network: ${network.id}
variables:
  default:
    fn::invoke:
      function: gcp:appengine:getDefaultServiceAccount
      arguments: {}

Setting privateInstance to true places the instance in a VPC with no public IP. The networkConfig property defines VPC peering with an allocated IP range. The enableStackdriverLogging and enableStackdriverMonitoring properties send logs and metrics to Cloud Logging and Monitoring. The dataprocServiceAccount grants permissions for Dataproc clusters that Data Fusion creates to run pipelines.

Encrypt instance data with customer-managed keys

Organizations with strict governance requirements use customer-managed encryption keys to control data encryption.

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

const keyRing = new gcp.kms.KeyRing("key_ring", {
    name: "my-instance",
    location: "us-central1",
});
const cryptoKey = new gcp.kms.CryptoKey("crypto_key", {
    name: "my-instance",
    keyRing: keyRing.id,
});
const project = gcp.organizations.getProject({});
const cryptoKeyMemberCdfSa = new gcp.kms.CryptoKeyIAMMember("crypto_key_member_cdf_sa", {
    cryptoKeyId: cryptoKey.id,
    role: "roles/cloudkms.cryptoKeyEncrypterDecrypter",
    member: project.then(project => `serviceAccount:service-${project.number}@gcp-sa-datafusion.iam.gserviceaccount.com`),
});
const cryptoKeyMemberGcsSa = new gcp.kms.CryptoKeyIAMMember("crypto_key_member_gcs_sa", {
    cryptoKeyId: cryptoKey.id,
    role: "roles/cloudkms.cryptoKeyEncrypterDecrypter",
    member: project.then(project => `serviceAccount:service-${project.number}@gs-project-accounts.iam.gserviceaccount.com`),
});
const cmek = new gcp.datafusion.Instance("cmek", {
    name: "my-instance",
    region: "us-central1",
    type: "BASIC",
    cryptoKeyConfig: {
        keyReference: cryptoKey.id,
    },
}, {
    dependsOn: [
        cryptoKeyMemberCdfSa,
        cryptoKeyMemberGcsSa,
    ],
});
import pulumi
import pulumi_gcp as gcp

key_ring = gcp.kms.KeyRing("key_ring",
    name="my-instance",
    location="us-central1")
crypto_key = gcp.kms.CryptoKey("crypto_key",
    name="my-instance",
    key_ring=key_ring.id)
project = gcp.organizations.get_project()
crypto_key_member_cdf_sa = gcp.kms.CryptoKeyIAMMember("crypto_key_member_cdf_sa",
    crypto_key_id=crypto_key.id,
    role="roles/cloudkms.cryptoKeyEncrypterDecrypter",
    member=f"serviceAccount:service-{project.number}@gcp-sa-datafusion.iam.gserviceaccount.com")
crypto_key_member_gcs_sa = gcp.kms.CryptoKeyIAMMember("crypto_key_member_gcs_sa",
    crypto_key_id=crypto_key.id,
    role="roles/cloudkms.cryptoKeyEncrypterDecrypter",
    member=f"serviceAccount:service-{project.number}@gs-project-accounts.iam.gserviceaccount.com")
cmek = gcp.datafusion.Instance("cmek",
    name="my-instance",
    region="us-central1",
    type="BASIC",
    crypto_key_config={
        "key_reference": crypto_key.id,
    },
    opts = pulumi.ResourceOptions(depends_on=[
            crypto_key_member_cdf_sa,
            crypto_key_member_gcs_sa,
        ]))
package main

import (
	"fmt"

	"github.com/pulumi/pulumi-gcp/sdk/v9/go/gcp/datafusion"
	"github.com/pulumi/pulumi-gcp/sdk/v9/go/gcp/kms"
	"github.com/pulumi/pulumi-gcp/sdk/v9/go/gcp/organizations"
	"github.com/pulumi/pulumi/sdk/v3/go/pulumi"
)

func main() {
	pulumi.Run(func(ctx *pulumi.Context) error {
		keyRing, err := kms.NewKeyRing(ctx, "key_ring", &kms.KeyRingArgs{
			Name:     pulumi.String("my-instance"),
			Location: pulumi.String("us-central1"),
		})
		if err != nil {
			return err
		}
		cryptoKey, err := kms.NewCryptoKey(ctx, "crypto_key", &kms.CryptoKeyArgs{
			Name:    pulumi.String("my-instance"),
			KeyRing: keyRing.ID(),
		})
		if err != nil {
			return err
		}
		project, err := organizations.LookupProject(ctx, &organizations.LookupProjectArgs{}, nil)
		if err != nil {
			return err
		}
		cryptoKeyMemberCdfSa, err := kms.NewCryptoKeyIAMMember(ctx, "crypto_key_member_cdf_sa", &kms.CryptoKeyIAMMemberArgs{
			CryptoKeyId: cryptoKey.ID(),
			Role:        pulumi.String("roles/cloudkms.cryptoKeyEncrypterDecrypter"),
			Member:      pulumi.Sprintf("serviceAccount:service-%v@gcp-sa-datafusion.iam.gserviceaccount.com", project.Number),
		})
		if err != nil {
			return err
		}
		cryptoKeyMemberGcsSa, err := kms.NewCryptoKeyIAMMember(ctx, "crypto_key_member_gcs_sa", &kms.CryptoKeyIAMMemberArgs{
			CryptoKeyId: cryptoKey.ID(),
			Role:        pulumi.String("roles/cloudkms.cryptoKeyEncrypterDecrypter"),
			Member:      pulumi.Sprintf("serviceAccount:service-%v@gs-project-accounts.iam.gserviceaccount.com", project.Number),
		})
		if err != nil {
			return err
		}
		_, err = datafusion.NewInstance(ctx, "cmek", &datafusion.InstanceArgs{
			Name:   pulumi.String("my-instance"),
			Region: pulumi.String("us-central1"),
			Type:   pulumi.String("BASIC"),
			CryptoKeyConfig: &datafusion.InstanceCryptoKeyConfigArgs{
				KeyReference: cryptoKey.ID(),
			},
		}, pulumi.DependsOn([]pulumi.Resource{
			cryptoKeyMemberCdfSa,
			cryptoKeyMemberGcsSa,
		}))
		if err != nil {
			return err
		}
		return nil
	})
}
using System.Collections.Generic;
using System.Linq;
using Pulumi;
using Gcp = Pulumi.Gcp;

return await Deployment.RunAsync(() => 
{
    var keyRing = new Gcp.Kms.KeyRing("key_ring", new()
    {
        Name = "my-instance",
        Location = "us-central1",
    });

    var cryptoKey = new Gcp.Kms.CryptoKey("crypto_key", new()
    {
        Name = "my-instance",
        KeyRing = keyRing.Id,
    });

    var project = Gcp.Organizations.GetProject.Invoke();

    var cryptoKeyMemberCdfSa = new Gcp.Kms.CryptoKeyIAMMember("crypto_key_member_cdf_sa", new()
    {
        CryptoKeyId = cryptoKey.Id,
        Role = "roles/cloudkms.cryptoKeyEncrypterDecrypter",
        Member = $"serviceAccount:service-{project.Apply(getProjectResult => getProjectResult.Number)}@gcp-sa-datafusion.iam.gserviceaccount.com",
    });

    var cryptoKeyMemberGcsSa = new Gcp.Kms.CryptoKeyIAMMember("crypto_key_member_gcs_sa", new()
    {
        CryptoKeyId = cryptoKey.Id,
        Role = "roles/cloudkms.cryptoKeyEncrypterDecrypter",
        Member = $"serviceAccount:service-{project.Apply(getProjectResult => getProjectResult.Number)}@gs-project-accounts.iam.gserviceaccount.com",
    });

    var cmek = new Gcp.DataFusion.Instance("cmek", new()
    {
        Name = "my-instance",
        Region = "us-central1",
        Type = "BASIC",
        CryptoKeyConfig = new Gcp.DataFusion.Inputs.InstanceCryptoKeyConfigArgs
        {
            KeyReference = cryptoKey.Id,
        },
    }, new CustomResourceOptions
    {
        DependsOn =
        {
            cryptoKeyMemberCdfSa,
            cryptoKeyMemberGcsSa,
        },
    });

});
package generated_program;

import com.pulumi.Context;
import com.pulumi.Pulumi;
import com.pulumi.core.Output;
import com.pulumi.gcp.kms.KeyRing;
import com.pulumi.gcp.kms.KeyRingArgs;
import com.pulumi.gcp.kms.CryptoKey;
import com.pulumi.gcp.kms.CryptoKeyArgs;
import com.pulumi.gcp.organizations.OrganizationsFunctions;
import com.pulumi.gcp.organizations.inputs.GetProjectArgs;
import com.pulumi.gcp.kms.CryptoKeyIAMMember;
import com.pulumi.gcp.kms.CryptoKeyIAMMemberArgs;
import com.pulumi.gcp.datafusion.Instance;
import com.pulumi.gcp.datafusion.InstanceArgs;
import com.pulumi.gcp.datafusion.inputs.InstanceCryptoKeyConfigArgs;
import com.pulumi.resources.CustomResourceOptions;
import java.util.List;
import java.util.ArrayList;
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 keyRing = new KeyRing("keyRing", KeyRingArgs.builder()
            .name("my-instance")
            .location("us-central1")
            .build());

        var cryptoKey = new CryptoKey("cryptoKey", CryptoKeyArgs.builder()
            .name("my-instance")
            .keyRing(keyRing.id())
            .build());

        final var project = OrganizationsFunctions.getProject(GetProjectArgs.builder()
            .build());

        var cryptoKeyMemberCdfSa = new CryptoKeyIAMMember("cryptoKeyMemberCdfSa", CryptoKeyIAMMemberArgs.builder()
            .cryptoKeyId(cryptoKey.id())
            .role("roles/cloudkms.cryptoKeyEncrypterDecrypter")
            .member(String.format("serviceAccount:service-%s@gcp-sa-datafusion.iam.gserviceaccount.com", project.number()))
            .build());

        var cryptoKeyMemberGcsSa = new CryptoKeyIAMMember("cryptoKeyMemberGcsSa", CryptoKeyIAMMemberArgs.builder()
            .cryptoKeyId(cryptoKey.id())
            .role("roles/cloudkms.cryptoKeyEncrypterDecrypter")
            .member(String.format("serviceAccount:service-%s@gs-project-accounts.iam.gserviceaccount.com", project.number()))
            .build());

        var cmek = new Instance("cmek", InstanceArgs.builder()
            .name("my-instance")
            .region("us-central1")
            .type("BASIC")
            .cryptoKeyConfig(InstanceCryptoKeyConfigArgs.builder()
                .keyReference(cryptoKey.id())
                .build())
            .build(), CustomResourceOptions.builder()
                .dependsOn(                
                    cryptoKeyMemberCdfSa,
                    cryptoKeyMemberGcsSa)
                .build());

    }
}
resources:
  cmek:
    type: gcp:datafusion:Instance
    properties:
      name: my-instance
      region: us-central1
      type: BASIC
      cryptoKeyConfig:
        keyReference: ${cryptoKey.id}
    options:
      dependsOn:
        - ${cryptoKeyMemberCdfSa}
        - ${cryptoKeyMemberGcsSa}
  cryptoKey:
    type: gcp:kms:CryptoKey
    name: crypto_key
    properties:
      name: my-instance
      keyRing: ${keyRing.id}
  keyRing:
    type: gcp:kms:KeyRing
    name: key_ring
    properties:
      name: my-instance
      location: us-central1
  cryptoKeyMemberCdfSa:
    type: gcp:kms:CryptoKeyIAMMember
    name: crypto_key_member_cdf_sa
    properties:
      cryptoKeyId: ${cryptoKey.id}
      role: roles/cloudkms.cryptoKeyEncrypterDecrypter
      member: serviceAccount:service-${project.number}@gcp-sa-datafusion.iam.gserviceaccount.com
  cryptoKeyMemberGcsSa:
    type: gcp:kms:CryptoKeyIAMMember
    name: crypto_key_member_gcs_sa
    properties:
      cryptoKeyId: ${cryptoKey.id}
      role: roles/cloudkms.cryptoKeyEncrypterDecrypter
      member: serviceAccount:service-${project.number}@gs-project-accounts.iam.gserviceaccount.com
variables:
  project:
    fn::invoke:
      function: gcp:organizations:getProject
      arguments: {}

The cryptoKeyConfig property references a Cloud KMS key for encrypting instance data and metadata. Before creating the instance, you must grant the Data Fusion and Cloud Storage service accounts cryptoKeyEncrypterDecrypter permissions on the key. The dependsOn ensures IAM bindings complete before instance creation.

Publish pipeline events to Pub/Sub

Data pipelines often notify downstream systems when jobs complete or fail.

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

const eventTopic = new gcp.pubsub.Topic("event", {name: "my-instance"});
const event = new gcp.datafusion.Instance("event", {
    name: "my-instance",
    region: "us-central1",
    type: "BASIC",
    eventPublishConfig: {
        enabled: true,
        topic: eventTopic.id,
    },
});
import pulumi
import pulumi_gcp as gcp

event_topic = gcp.pubsub.Topic("event", name="my-instance")
event = gcp.datafusion.Instance("event",
    name="my-instance",
    region="us-central1",
    type="BASIC",
    event_publish_config={
        "enabled": True,
        "topic": event_topic.id,
    })
package main

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

func main() {
	pulumi.Run(func(ctx *pulumi.Context) error {
		eventTopic, err := pubsub.NewTopic(ctx, "event", &pubsub.TopicArgs{
			Name: pulumi.String("my-instance"),
		})
		if err != nil {
			return err
		}
		_, err = datafusion.NewInstance(ctx, "event", &datafusion.InstanceArgs{
			Name:   pulumi.String("my-instance"),
			Region: pulumi.String("us-central1"),
			Type:   pulumi.String("BASIC"),
			EventPublishConfig: &datafusion.InstanceEventPublishConfigArgs{
				Enabled: pulumi.Bool(true),
				Topic:   eventTopic.ID(),
			},
		})
		if err != nil {
			return err
		}
		return nil
	})
}
using System.Collections.Generic;
using System.Linq;
using Pulumi;
using Gcp = Pulumi.Gcp;

return await Deployment.RunAsync(() => 
{
    var eventTopic = new Gcp.PubSub.Topic("event", new()
    {
        Name = "my-instance",
    });

    var @event = new Gcp.DataFusion.Instance("event", new()
    {
        Name = "my-instance",
        Region = "us-central1",
        Type = "BASIC",
        EventPublishConfig = new Gcp.DataFusion.Inputs.InstanceEventPublishConfigArgs
        {
            Enabled = true,
            Topic = eventTopic.Id,
        },
    });

});
package generated_program;

import com.pulumi.Context;
import com.pulumi.Pulumi;
import com.pulumi.core.Output;
import com.pulumi.gcp.pubsub.Topic;
import com.pulumi.gcp.pubsub.TopicArgs;
import com.pulumi.gcp.datafusion.Instance;
import com.pulumi.gcp.datafusion.InstanceArgs;
import com.pulumi.gcp.datafusion.inputs.InstanceEventPublishConfigArgs;
import java.util.List;
import java.util.ArrayList;
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 eventTopic = new Topic("eventTopic", TopicArgs.builder()
            .name("my-instance")
            .build());

        var event = new Instance("event", InstanceArgs.builder()
            .name("my-instance")
            .region("us-central1")
            .type("BASIC")
            .eventPublishConfig(InstanceEventPublishConfigArgs.builder()
                .enabled(true)
                .topic(eventTopic.id())
                .build())
            .build());

    }
}
resources:
  event:
    type: gcp:datafusion:Instance
    properties:
      name: my-instance
      region: us-central1
      type: BASIC
      eventPublishConfig:
        enabled: true
        topic: ${eventTopic.id}
  eventTopic:
    type: gcp:pubsub:Topic
    name: event
    properties:
      name: my-instance

The eventPublishConfig property enables pipeline event publishing and specifies the destination Pub/Sub topic. When enabled, Data Fusion sends lifecycle events (pipeline start, success, failure) to the topic, allowing external systems to react to pipeline state changes.

Beyond these examples

These snippets focus on specific Data Fusion instance features: instance types and tiers, private networking and VPC configuration, and customer-managed encryption keys. They’re intentionally minimal rather than full data integration platforms.

The examples may reference pre-existing infrastructure such as VPC networks, subnets, and IP allocations for private instances, KMS keys and IAM bindings for CMEK, Pub/Sub topics for event publishing, and service accounts. They focus on configuring the instance rather than provisioning everything around it.

To keep things focused, common instance patterns are omitted, including:

  • Accelerator configuration (CDC, other types)
  • RBAC enablement for Enterprise tier (enableRbac)
  • Zone-specific placement (zone property)
  • Private Service Connect configuration
  • Display names and labels for organization

These omissions are intentional: the goal is to illustrate how each Data Fusion feature is wired, not provide drop-in data platform modules. See the Data Fusion Instance resource reference for all available configuration options.

Let's manage GCP Data Fusion Instance IAM Bindings

Get started with Pulumi Cloud, then follow our quick setup guide to deploy this infrastructure.

Try Pulumi Cloud for FREE

Frequently Asked Questions

Configuration & Immutability
What properties can't be changed after creating an IAM binding?
The name, project, region, role, and condition properties are immutable. Only members can be modified after creation.
Can I change the region or role after creation?
No, both region and role are immutable. You’ll need to destroy and recreate the binding to change these properties.
What instance types are available?
The examples show three types: BASIC (standard instances), ENTERPRISE (with RBAC support), and DEVELOPER (for development/testing).
Encryption & Security
How do I set up customer-managed encryption keys (CMEK) for my instance?
Configure cryptoKeyConfig with a keyReference pointing to your KMS key. Use dependsOn to ensure the Data Fusion and GCS service accounts have roles/cloudkms.cryptoKeyEncrypterDecrypter permissions before creating the instance.
Networking
How do I configure Private Service Connect for my Data Fusion instance?
Set networkConfig.connectionType to PRIVATE_SERVICE_CONNECT_INTERFACES and provide privateServiceConnectConfig with a networkAttachment and unreachableCidrBlock (e.g., 192.168.0.0/25).
Import & Management
What formats can I use to import an existing instance IAM binding?
You can use four formats: projects/{{project}}/locations/{{region}}/instances/{{name}}, {{project}}/{{region}}/{{name}}, {{region}}/{{name}}, or just {{name}}.

Using a different cloud?

Explore security guides for other cloud providers: