Create instance templates

This page describes how to create and manage instance templates. Instance templates let you specify the machine type, boot disk image, network, and other VM properties that you want to use when creating virtual machine (VM) instances.

You can use instance templates to do the following:

Before you begin

  • Read when and why to create deterministic instance templates.
  • Read about regional and global instance templates.
  • If you haven't already, then set up authentication. Authentication is the process by which your identity is verified for access to Google Cloud services and APIs. To run code or samples from a local development environment, you can authenticate to Compute Engine by selecting one of the following options:

    Select the tab for how you plan to use the samples on this page:

    Console

    When you use the Google Cloud console to access Google Cloud services and APIs, you don't need to set up authentication.

    gcloud

    1. Install the Google Cloud CLI, then initialize it by running the following command: 

      gcloud init
    2. Set a default region and zone.

    Terraform

    To use the Terraform samples on this page in a local development environment, install and initialize the gcloud CLI, and then set up Application Default Credentials with your user credentials.

    1. Install the Google Cloud CLI.

    2. To initialize the gcloud CLI, run the following command: 

      gcloud init

    3. If you're using a local shell, then create local authentication credentials for your user account: 

      gcloud auth application-default login

      You don't need to do this if you're using Cloud Shell.

    For more information, see Set up authentication for a local development environment.

    Go

    To use the Go samples on this page in a local development environment, install and initialize the gcloud CLI, and then set up Application Default Credentials with your user credentials.

    1. Install the Google Cloud CLI.

    2. To initialize the gcloud CLI, run the following command: 

      gcloud init

    3. If you're using a local shell, then create local authentication credentials for your user account: 

      gcloud auth application-default login

      You don't need to do this if you're using Cloud Shell.

    For more information, see Set up authentication for a local development environment.

    Java

    To use the Java samples on this page in a local development environment, install and initialize the gcloud CLI, and then set up Application Default Credentials with your user credentials.

    1. Install the Google Cloud CLI.

    2. To initialize the gcloud CLI, run the following command: 

      gcloud init

    3. If you're using a local shell, then create local authentication credentials for your user account: 

      gcloud auth application-default login

      You don't need to do this if you're using Cloud Shell.

    For more information, see Set up authentication for a local development environment.

    Node.js

    To use the Node.js samples on this page in a local development environment, install and initialize the gcloud CLI, and then set up Application Default Credentials with your user credentials.

    1. Install the Google Cloud CLI.

    2. To initialize the gcloud CLI, run the following command: 

      gcloud init

    3. If you're using a local shell, then create local authentication credentials for your user account: 

      gcloud auth application-default login

      You don't need to do this if you're using Cloud Shell.

    For more information, see Set up authentication for a local development environment.

    Python

    To use the Python samples on this page in a local development environment, install and initialize the gcloud CLI, and then set up Application Default Credentials with your user credentials.

    1. Install the Google Cloud CLI.

    2. To initialize the gcloud CLI, run the following command: 

      gcloud init

    3. If you're using a local shell, then create local authentication credentials for your user account: 

      gcloud auth application-default login

      You don't need to do this if you're using Cloud Shell.

    For more information, see Set up authentication for a local development environment.

    REST

    To use the REST API samples on this page in a local development environment, you use the credentials you provide to the gcloud CLI.

      Install the Google Cloud CLI, then initialize it by running the following command: 

      gcloud init

    For more information, see Authenticate for using REST in the Google Cloud authentication documentation.

Limitations

  • Shared VPC on interfaces other than nic0 for instance templates is supported in gcloud CLI and REST, but not in Google Cloud console.
  • You can't update an existing instance template or change an instance template after it has been created. If an instance template goes out of date, or you need to make changes to the configuration, create a new instance template.
  • If you want to specify an image family in an instance template, you can't use the Google Cloud console. You can use the Google Cloud CLI or REST instead.
  • If you want to specify regional disks instead of zonal disks in an instance template, you can't use the Google Cloud console. You can use the Google Cloud CLI or REST instead.
  • You can use an instance template to create VMs with a Hyperdisk Balanced boot disk that is in a storage pool, if the storage pool exists in the same zone that the VM is created in. You can't use global instance templates to create VMs with non-boot disks that are in a storage pool.

Create an instance template

Most of the VM properties that you can specify in a request to create an individual VM instance can also be specified for an instance template, including any VM metadata, startup scripts, persistent disks, and service accounts. You must specify the machine type, boot disk, and network.

Create a regional or global instance template through the Google Cloud console, Google Cloud CLI, or the API. To create a global instance template, you can also use the Terraform or the Cloud Client Libraries.

Console

  1. In the Google Cloud console, go to the Instance templates page.

    Go to Instance templates

    The remaining steps appear in the Google Cloud console.

  2. Click Create instance template.
  3. Select the Location as follows:
    1. If you want to use the instance template across regions, choose Global.
    2. If you want to reduce cross-region dependency, choose Regional.
  4. If you chose regional, then select the Region where you want to create your instance template.

  5. For the following fields, either accept the default values or modify them as required. The default values change based on the machine family that you select.

    • Select a Machine type.
    • To update the boot disk type or image, in the Boot disk section, click Change.
    • To update the network interface or IP address settings, click Advanced options, then click Networking, and then click the network interface you want to edit.

  6. Optional: If you chose an image that supports Shielded VM, change the VM's Shielded VM settings:

    1. Click Advanced options, and then click the Security tab.
    2. If you want to disable Secure Boot, clear the Turn on Secure Boot checkbox. Secure Boot helps protect your VM instances against boot-level and kernel-level malware and rootkits. For more information, see Secure boot.

    3. If you want to disable the virtual trusted platform module (vTPM), clear the Turn on vTPM checkbox. The vTPM enables Measured boot, which validates the VM pre-boot and boot integrity. For more information, see Virtual Trusted Platform Module (vTPM).

    4. If you want to disable integrity monitoring, uncheck the Turn on Integrity Monitoring checkbox. Integrity monitoring lets youn monitor the boot integrity of your Shielded VM instances by using Cloud Monitoring. For more information, see Integrity monitoring.

  7. Optional: Under Advanced options, click the tabs to further customize your template. For example, you can add up to 15 secondary non-boot disks.

  8. Optional: Click Equivalent REST to view the REST request body, which includes the JSON representation of your instance template.

  9. Click Create to create the template.

gcloud

To create a regional or global instance template, use the instance-templates create command. For a regional instance template, you must use the --instance-template-region flag to set the region of the template.

Create a regional instance template using the following command.

gcloud compute instance-templates create INSTANCE_TEMPLATE_NAME \
    --instance-template-region=REGION

Replace REGION with the region where you want to create the regional instance template.

Create a global instance template using the following command:

gcloud compute instance-templates create INSTANCE_TEMPLATE_NAME

If you don't provide explicit template settings, gcloud compute uses the following default values:

  • Machine type: the machine type—for example, n1-standard-1
  • Image: the latest Debian image
  • Boot disk: a new standard boot disk named after the VM
  • Network: the default VPC network
  • IP address: an ephemeral external IPv4 address
  • Stack type: IPV4_ONLY

You can also explicitly provide these configuration settings. For example:

gcloud compute instance-templates create example-template-custom \
    --machine-type=e2-standard-4 \
    --image-family=debian-12 \
    --image-project=debian-cloud \
    --boot-disk-size=250GB

You can add up to 15 secondary non-boot disks. Specify the --create-disk flag for each secondary disk you create. To create secondary disks from a public or custom image, specify the image and image-project properties for each disk in the --create-disk flag. To create a blank disk, don't include these properties. Optionally, include properties for the disk size and type. To specify regional disks, use the replica-zones property.

gcloud compute instance-templates create INSTANCE_TEMPLATE_NAME \
    --machine-type=MACHINE_TYPE \
    --create-disk=image-family=DISK_IMAGE_FAMILY,image-project=DISK_IMAGE_PROJECT,size=SIZE_GB_DISK1 \
    --create-disk=device-name=DISK_NAME,type=DISK_TYPE,size=SIZE_GB_DISK2,replica-zones=^:^ZONE:REMOTE_ZONE,boot=false

Tip: When specifying the disk replica-zones parameter, the characters ^:^ specify that the separation character between values is a colon (:) instead of the expected comma (,).

Replace the following:

  • INSTANCE_TEMPLATE_NAME: the name for the template
  • MACHINE_TYPE: the machine type of the VMs

  • DISK_IMAGE_FAMILY: an image family to use as a non-boot disk

    For more information about image families, see best practices when using image families on Compute Engine.

    You can use instead the flag --image=IMAGE to specify a specific version of an image.

    For blank disks, don't specify the image-family or image property.

  • DISK_IMAGE_PROJECT: the image project that contains the image

    For blank disks, don't specify the image-project property. For more information on public images, see Public images.

  • SIZE_GB_DISK1 and SIZE_GB_DISK2: the size of each secondary disk

  • DISK_NAME: Optional: the disk name displayed to the guest OS after the VM is created.

  • DISK_TYPE: Optional: the type of disk to create. If not specified, the default disk type that is used depends on the value of the --machine-type flag.

  • ZONE and REMOTE_ZONE: the zone to create the regional disk in and the zone to replicate it to.

    For zonal disks, don't include the replica-zones property.

If you chose an image that supports Shielded VM, you can optionally change the instance's Shielded VM settings using one of the following flags:

  • --no-shielded-secure-boot: turns off Secure Boot

    Secure Boot helps protect your VM instances against boot-level and kernel-level malware and rootkits. For more information, see Secure Boot.

  • --no-shielded-vtpm: turns off the virtual trusted platform module (vTPM)

    The vTPM enables Measured Boot, which validates the VM pre-boot and boot integrity. For more information, see Virtual Trusted Platform Module (vTPM).

  • --no-shielded-integrity-monitoring: turns off integrity monitoring

    Integrity monitoring lets you monitor the boot integrity of your Shielded VM instances using Cloud Monitoring. For more information, see Integrity monitoring.

For a list of all available subcommands and flags, see the instance-templates reference.

A template with the default configuration settings might look like the following:

gcloud compute instance-templates describe example-template

creationTimestamp: '2019-09-10T16:18:32.042-07:00'
description: ''
id: '6057583701980539406'
kind: compute#instanceTemplate
name: example-template
properties:
  canIpForward: false
  disks:
  - autoDelete: true
    boot: true
    initializeParams:
      sourceImage: https://compute.googleapis.com/compute/v1/projects/debian-cloud/global/images/family/debian-12
    kind: compute#attachedDisk
    mode: READ_WRITE
    type: PERSISTENT
  machineType: c3-standard-4
  networkInterfaces:
  - accessConfigs:
    - kind: compute#accessConfig
      name: external-nat
      type: ONE_TO_ONE_NAT
    network: https://compute.googleapis.com/compute/v1/projects/myproject/global/networks/default
  scheduling:
    automaticRestart: true
    onHostMaintenance: MIGRATE
  serviceAccounts:
  - email: default
    scopes:
    - https://www.googleapis.com/auth/devstorage.read_only
selfLink: https://compute.googleapis.com/compute/v1/projects/myproject/global/instanceTemplates/example-template

Terraform

To create an instance template, you can use the google_compute_instance_template resource.

The following Terraform example is similar to the following gcloud CLI command:

gcloud compute instance-templates create my-instance-template \
    --machine-type=c3-standard-4 \
    --image-family=debian-12 \
    --image-project=debian-cloud \
    --boot-disk-size=250GB

resource "google_compute_instance_template" "foobar" {
  name         = "my-instance-template"
  machine_type = "e2-standard-4"

  disk {
    source_image = "debian-cloud/debian-11"
    disk_size_gb = 250
  }

  network_interface {
    network = "default"

    # default access config, defining external IP configuration
    access_config {
      network_tier = "PREMIUM"
    }
  }

  # To avoid embedding secret keys or user credentials in the instances, Google recommends that you use custom service accounts with the following access scopes.
  service_account {
    scopes = [
      "https://www.googleapis.com/auth/cloud-platform"
    ]
  }
}

To learn how to apply or remove a Terraform configuration, see Basic Terraform commands.

Go 

import (
	"context"
	"fmt"
	"io"

	compute "cloud.google.com/go/compute/apiv1"
	computepb "cloud.google.com/go/compute/apiv1/computepb"
	"google.golang.org/protobuf/proto"
)

// createTemplate creates a new instance template with the provided name and a specific instance configuration.
func createTemplate(w io.Writer, projectID, templateName string) error {
	// projectID := "your_project_id"
	// templateName := "your_template_name"

	ctx := context.Background()
	instanceTemplatesClient, err := compute.NewInstanceTemplatesRESTClient(ctx)
	if err != nil {
		return fmt.Errorf("NewInstanceTemplatesRESTClient: %w", err)
	}
	defer instanceTemplatesClient.Close()

	req := &computepb.InsertInstanceTemplateRequest{
		Project: projectID,
		InstanceTemplateResource: &computepb.InstanceTemplate{
			Name: proto.String(templateName),
			Properties: &computepb.InstanceProperties{
				// The template describes the size and source image of the boot disk
				// to attach to the instance.
				Disks: []*computepb.AttachedDisk{
					{
						InitializeParams: &computepb.AttachedDiskInitializeParams{
							DiskSizeGb:  proto.Int64(250),
							SourceImage: proto.String("projects/debian-cloud/global/images/family/debian-11"),
						},
						AutoDelete: proto.Bool(true),
						Boot:       proto.Bool(true),
					},
				},
				MachineType: proto.String("e2-standard-4"),
				// The template connects the instance to the `default` network,
				// without specifying a subnetwork.
				NetworkInterfaces: []*computepb.NetworkInterface{
					{
						Name: proto.String("global/networks/default"),
						// The template lets the instance use an external IP address.
						AccessConfigs: []*computepb.AccessConfig{
							{
								Name:        proto.String("External NAT"),
								Type:        proto.String(computepb.AccessConfig_ONE_TO_ONE_NAT.String()),
								NetworkTier: proto.String(computepb.AccessConfig_PREMIUM.String()),
							},
						},
					},
				},
			},
		},
	}

	op, err := instanceTemplatesClient.Insert(ctx, req)
	if err != nil {
		return fmt.Errorf("unable to create instance template: %w", err)
	}

	if err = op.Wait(ctx); err != nil {
		return fmt.Errorf("unable to wait for the operation: %w", err)
	}

	fmt.Fprintf(w, "Instance template created\n")

	return nil
}

Java 

import com.google.cloud.compute.v1.AccessConfig;
import com.google.cloud.compute.v1.AccessConfig.NetworkTier;
import com.google.cloud.compute.v1.AttachedDisk;
import com.google.cloud.compute.v1.AttachedDiskInitializeParams;
import com.google.cloud.compute.v1.GlobalOperationsClient;
import com.google.cloud.compute.v1.InsertInstanceTemplateRequest;
import com.google.cloud.compute.v1.InstanceProperties;
import com.google.cloud.compute.v1.InstanceTemplate;
import com.google.cloud.compute.v1.InstanceTemplatesClient;
import com.google.cloud.compute.v1.NetworkInterface;
import com.google.cloud.compute.v1.Operation;
import java.io.IOException;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;

public class CreateInstanceTemplate {

  public static void main(String[] args)
      throws IOException, ExecutionException, InterruptedException, TimeoutException {
    // TODO(developer): Replace these variables before running the sample.
    // projectId: project ID or project number of the Cloud project you use.
    // templateName: name of the new template to create.
    String projectId = "your-project-id";
    String templateName = "template-name";
    createInstanceTemplate(projectId, templateName);
  }

  /*
    Create a new instance template with the provided name and a specific
    instance configuration.
   */
  public static void createInstanceTemplate(String projectId, String templateName)
      throws IOException, ExecutionException, InterruptedException, TimeoutException {
    try (InstanceTemplatesClient instanceTemplatesClient = InstanceTemplatesClient.create()) {

      String machineType = "e2-standard-4";
      String sourceImage = "projects/debian-cloud/global/images/family/debian-11";

      // The template describes the size and source image of the boot disk
      // to attach to the instance.
      AttachedDisk attachedDisk = AttachedDisk.newBuilder()
          .setInitializeParams(AttachedDiskInitializeParams.newBuilder()
              .setSourceImage(sourceImage)
              .setDiskType("pd-balanced")
              .setDiskSizeGb(250).build())
          .setAutoDelete(true)
          .setBoot(true).build();

      // The template connects the instance to the `default` network,
      // without specifying a subnetwork.
      NetworkInterface networkInterface = NetworkInterface.newBuilder()
          .setName("global/networks/default")
          // The template lets the instance use an external IP address.
          .addAccessConfigs(AccessConfig.newBuilder()
              .setName("External NAT")
              .setType(AccessConfig.Type.ONE_TO_ONE_NAT.toString())
              .setNetworkTier(NetworkTier.PREMIUM.toString()).build()).build();

      InstanceProperties instanceProperties = InstanceProperties.newBuilder()
          .addDisks(attachedDisk)
          .setMachineType(machineType)
          .addNetworkInterfaces(networkInterface).build();

      InsertInstanceTemplateRequest insertInstanceTemplateRequest = InsertInstanceTemplateRequest
          .newBuilder()
          .setProject(projectId)
          .setInstanceTemplateResource(InstanceTemplate.newBuilder()
              .setName(templateName)
              .setProperties(instanceProperties).build()).build();

      // Create the Instance Template.
      Operation response = instanceTemplatesClient.insertAsync(insertInstanceTemplateRequest)
          .get(3, TimeUnit.MINUTES);

      if (response.hasError()) {
        System.out.println("Instance Template creation failed ! ! " + response);
        return;
      }
      System.out
          .printf("Instance Template Operation Status %s: %s", templateName, response.getStatus());
    }
  }

  public static void createInstanceTemplateWithDiskType(String projectId, String templateName)
      throws IOException, ExecutionException, InterruptedException, TimeoutException {
    try (InstanceTemplatesClient instanceTemplatesClient = InstanceTemplatesClient.create();
        GlobalOperationsClient globalOperationsClient = GlobalOperationsClient.create()) {

      AttachedDisk disk = AttachedDisk.newBuilder()
          .setInitializeParams(AttachedDiskInitializeParams.newBuilder()
              .setDiskSizeGb(10)
              .setDiskType("pd-balanced")
              .setSourceImage("projects/debian-cloud/global/images/family/debian-11").build())
          .setAutoDelete(true)
          .setBoot(true)
          .setType(AttachedDisk.Type.PERSISTENT.toString()).build();

      InstanceTemplate instanceTemplate = InstanceTemplate.newBuilder()
          .setName(templateName)
          .setProperties(InstanceProperties.newBuilder()
              .setMachineType("n1-standard-1")
              .addDisks(disk)
              .addNetworkInterfaces(NetworkInterface.newBuilder()
                  .setName("global/networks/default").build()).build()).build();

      InsertInstanceTemplateRequest insertInstanceTemplateRequest = InsertInstanceTemplateRequest
          .newBuilder()
          .setProject(projectId)
          .setInstanceTemplateResource(instanceTemplate).build();

      Operation response = instanceTemplatesClient.insertAsync(insertInstanceTemplateRequest)
          .get(3, TimeUnit.MINUTES);

      if (response.hasError()) {
        System.out.println("Instance Template creation failed ! ! " + response);
        return;
      }
      System.out
          .printf("Instance Template Operation Status %s: %s", templateName, response.getStatus());
    }
  }
}

Node.js 

/**
 * TODO(developer): Uncomment and replace these variables before running the sample.
 */
// const projectId = 'YOUR_PROJECT_ID';
// const templateName = 'your_template_name';

const compute = require('@google-cloud/compute');

// Create a new instance template with the provided name and a specific instance configuration.
async function createTemplate() {
  const instanceTemplatesClient = new compute.InstanceTemplatesClient();

  const [response] = await instanceTemplatesClient.insert({
    project: projectId,
    instanceTemplateResource: {
      name: templateName,
      properties: {
        disks: [
          {
            // The template describes the size and source image of the boot disk
            // to attach to the instance.
            initializeParams: {
              diskSizeGb: '250',
              sourceImage:
                'projects/debian-cloud/global/images/family/debian-11',
            },
            autoDelete: true,
            boot: true,
          },
        ],
        machineType: 'e2-standard-4',
        // The template connects the instance to the `default` network,
        // without specifying a subnetwork.
        networkInterfaces: [
          {
            // Use the network interface provided in the networkName argument.
            name: 'global/networks/default',
            // The template lets the instance use an external IP address.
            accessConfigs: [
              {
                name: 'External NAT',
                type: 'ONE_TO_ONE_NAT',
                networkTier: 'PREMIUM',
              },
            ],
          },
        ],
      },
    },
  });
  let operation = response.latestResponse;
  const operationsClient = new compute.GlobalOperationsClient();

  // Wait for the create operation to complete.
  while (operation.status !== 'DONE') {
    [operation] = await operationsClient.wait({
      operation: operation.name,
      project: projectId,
    });
  }

  console.log('Instance template created.');
}

createTemplate();

Python 

from __future__ import annotations

import sys
from typing import Any

from google.api_core.extended_operation import ExtendedOperation
from google.cloud import compute_v1


def wait_for_extended_operation(
    operation: ExtendedOperation, verbose_name: str = "operation", timeout: int = 300
) -> Any:
    """
    Waits for the extended (long-running) operation to complete.

    If the operation is successful, it will return its result.
    If the operation ends with an error, an exception will be raised.
    If there were any warnings during the execution of the operation
    they will be printed to sys.stderr.

    Args:
        operation: a long-running operation you want to wait on.
        verbose_name: (optional) a more verbose name of the operation,
            used only during error and warning reporting.
        timeout: how long (in seconds) to wait for operation to finish.
            If None, wait indefinitely.

    Returns:
        Whatever the operation.result() returns.

    Raises:
        This method will raise the exception received from `operation.exception()`
        or RuntimeError if there is no exception set, but there is an `error_code`
        set for the `operation`.

        In case of an operation taking longer than `timeout` seconds to complete,
        a `concurrent.futures.TimeoutError` will be raised.
    """
    result = operation.result(timeout=timeout)

    if operation.error_code:
        print(
            f"Error during {verbose_name}: [Code: {operation.error_code}]: {operation.error_message}",
            file=sys.stderr,
            flush=True,
        )
        print(f"Operation ID: {operation.name}", file=sys.stderr, flush=True)
        raise operation.exception() or RuntimeError(operation.error_message)

    if operation.warnings:
        print(f"Warnings during {verbose_name}:\n", file=sys.stderr, flush=True)
        for warning in operation.warnings:
            print(f" - {warning.code}: {warning.message}", file=sys.stderr, flush=True)

    return result


def create_template(project_id: str, template_name: str) -> compute_v1.InstanceTemplate:
    """
    Create a new instance template with the provided name and a specific
    instance configuration.

    Args:
        project_id: project ID or project number of the Cloud project you use.
        template_name: name of the new template to create.

    Returns:
        InstanceTemplate object that represents the new instance template.
    """
    # The template describes the size and source image of the boot disk
    # to attach to the instance.
    disk = compute_v1.AttachedDisk()
    initialize_params = compute_v1.AttachedDiskInitializeParams()
    initialize_params.source_image = (
        "projects/debian-cloud/global/images/family/debian-11"
    )
    initialize_params.disk_size_gb = 250
    disk.initialize_params = initialize_params
    disk.auto_delete = True
    disk.boot = True

    # The template connects the instance to the `default` network,
    # without specifying a subnetwork.
    network_interface = compute_v1.NetworkInterface()
    network_interface.name = "global/networks/default"

    # The template lets the instance use an external IP address.
    access_config = compute_v1.AccessConfig()
    access_config.name = "External NAT"
    access_config.type_ = "ONE_TO_ONE_NAT"
    access_config.network_tier = "PREMIUM"
    network_interface.access_configs = [access_config]

    template = compute_v1.InstanceTemplate()
    template.name = template_name
    template.properties.disks = [disk]
    template.properties.machine_type = "e2-standard-4"
    template.properties.network_interfaces = [network_interface]

    template_client = compute_v1.InstanceTemplatesClient()
    operation = template_client.insert(
        project=project_id, instance_template_resource=template
    )

    wait_for_extended_operation(operation, "instance template creation")

    return template_client.get(project=project_id, instance_template=template_name)

REST

To create a regional instance template, make a POST request to the regionInstanceTemplates.insert method as follows:

POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/regions/REGION/instanceTemplates

To create a global instance template, make a POST request to the instanceTemplates.insert method:

POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/global/instanceTemplates

You can add up to 15 secondary non-boot disks by using the disks property, with a field for each additional disk. For each additional disk, you can do the following:

  • Create additional disks with a public or custom image.
  • To add a blank disk, define the initializeParams entry with no sourceImage value.
  • To create regional disks, define the initializeParams entry with the property replicaZones.

In the body of the request, provide the template properties:

{
  "name": "INSTANCE_TEMPLATE_NAME",
  "properties": {
    "machineType": "MACHINE_TYPE",
    "networkInterfaces": [
      {
        "network": "global/networks/default",
        "accessConfigs":
        [
          {
            "name": "external-IP",
            "type": "ONE_TO_ONE_NAT"
          }
        ]
      }
    ],
    "disks":
    [
      {
        "type": "PERSISTENT",
        "boot": true,
        "mode": "READ_WRITE",
        "initializeParams":
        {
          "sourceImage": "projects/IMAGE_PROJECT/global/images/IMAGE"
        }
      },
      {
        "type": "PERSISTENT",
        "boot": false,
        "deviceName": "DISK_NAME",
        "initializeParams":
        {
          "replicaZones": [
              "projects/PROJECT_NAME/zones/ZONE",
              "projects/PROJECT_NAME/zones/REMOTE_ZONE"
          ]
        }
      }
    ]
  }
}

Replace the following:

  • PROJECT_ID: your project id
  • REGION: the region where you want to create your regional instance template
  • INSTANCE_TEMPLATE_NAME: the name of the instance template
  • ZONE: the zone where VMs are located

  • MACHINE_TYPE: the machine type of the VMs

  • IMAGE_PROJECT: the image project that contains the image

    For more information on public images, see Public images.

  • IMAGE or IMAGE_FAMILY: specify one of the following:
    • IMAGE: a specific version of the image

      For example, "sourceImage": "projects/debian-cloud/global/images/debian-10-buster-v20200309"

    • IMAGE_FAMILY: an image family

      This creates the VM from the most recent, non-deprecated OS image. For example, if you specify "sourceImage": "projects/debian-cloud/global/images/family/debian-10", Compute Engine creates a VM from the latest version of the OS image in the Debian 10 image family.

      For more information on image families, see best practices when using image families on Compute Engine.

  • DISK_NAME: Optional: the disk name displayed to the guest OS after the VM is created.

  • PROJECT_NAME: the project associated with the VM

  • REMOTE_ZONE: the zone where the regional disk should be replicated to

You can specify one of the following options for the disks property:

  • Specify initializeParams to create boot disks for each instance. You can create disks by using public or custom images (or image families) using the sourceImage property, as shown in the preceding example. To add blank disks, don't specify a sourceImage. You can also add up to 15 secondary non-boot disks by using the initializeParams property for each additional disk.

  • Specify source to attach an existing boot disk. If you attach an existing boot disk, you can only create one instance from your template.

Optionally, you can specify the diskSizeGb, diskType, and labels properties for initializeParams and the diskSizeGb property for source.

If you chose an image that supports Shielded VM, you can optionally change the VM's Shielded VM settings by using the following Boolean request body items:

  • enableSecureBoot: turns on or off Secure Boot

    Secure Boot helps protect your VM instances against boot-level and kernel-level malware and rootkits. For more information, see Secure Boot.

  • enableVtpm: turns on or off the virtual trusted platform module (vTPM)

    The vTPM enables Measured Boot, which validates the VM pre-boot and boot integrity. For more information, see Virtual Trusted Platform Module (vTPM).

  • enableIntegrityMonitoring: turns on or off integrity monitoring

    Integrity monitoring lets you monitor and verify the runtime boot integrity of your Shielded VM instances by using Cloud Monitoring reports. For more information, see Integrity monitoring.

To learn more about request parameters, see the instanceTemplates.insert method.

Create an instance template based on an existing instance

You can use REST or gcloud CLI to save the configuration of an existing VM instance as an instance template. You can optionally override how the source disks are defined in the template.

If you need to override other properties, first create an instance template based on an existing instance, then create a similar template with additional overrides.

gcloud

Use the gcloud compute instance-templates create command with the --source-instance and --source-instance-zone flags. If you want to create a regional instance template, then you must also use the --instance-template-region flag to specify the region of the instance template.

To create a regional instance template, use the following command:

 gcloud compute instance-templates create INSTANCE_TEMPLATE_NAME \
     --source-instance=SOURCE_INSTANCE \
     --source-instance-zone=SOURCE_INSTANCE_ZONE \
     --instance-template-region=REGION

To create a global instance template, use the following command:

gcloud compute instance-templates create INSTANCE_TEMPLATE_NAME \
    --source-instance=SOURCE_INSTANCE \
    --source-instance-zone=SOURCE_INSTANCE_ZONE

To override how the source instance's disks are defined, add one or more --configure-disk flags.

The following example creates a global instance template from an existing instance and overrides the source instance's disk with the specifications that you provide.

  gcloud compute instance-templates create INSTANCE_TEMPLATE_NAME \
      --source-instance=SOURCE_INSTANCE \
      --source-instance-zone=SOURCE_INSTANCE_ZONE \
      --configure-disk= \
      device-name=SOURCE_DISK, \
      instantiate-from=INSTANTIATE_OPTIONS, \
      auto-delete=AUTO_DELETE

Replace the following:

  • INSTANCE_TEMPLATE_NAME is the name of the template to create.
  • SOURCE_INSTANCE is the name of the instance to use as a model for the new template.
  • SOURCE_INSTANCE_ZONE is the zone that contains the source instance.
  • SOURCE_DISK is the name of a source-instance disk that you want to override within the template.

  • INSTANTIATE_OPTIONS specifies whether to include the disk and which image to use. Valid values depend on the type of disk:

    • source-image or source-image-family (valid only for boot and other persistent read-write disks). Specify this option if you want to use the same source image or source image family that was used to create the disk in the source VM instance.

    • custom-image (valid only for boot and other persistent read-write disks). If you want to retain applications and settings from the source VMs in your instance template, you can create a custom image and then specify it when you create the template. If specified, then provide the path or URL for the custom image, as shown in the following example. Alternatively, you can specify an image family using the following format:

      --configure-disk=device-name=DATA_DISK_NAME,instantiate-from=custom-image, \
custom-image=projects/PROJECT_ID/global/images/family/IMAGE_FAMILY_NAME

    • attach-read-only (valid only for read-only disks).

    • blank (valid only for non-boot persistent disks and local SSDs). If specified, then, when the template is used to create a new instance, the disk is created unformatted. You must format and mount the disk in a startup script before you can use it in a scalable setup.

    • do-not-include (valid only for non-boot persistent disks and read-only disks).

  • AUTO_DELETE specifies whether the disk is auto-deleted when the instance is deleted. Valid values are: false, no, true, and yes.

For example, the following command creates an instance template based on my-source-instance, with the option to use the original image from data-disk-a, but set auto-delete to true and replace data-disk-b with a custom image.

 gcloud compute instance-templates create my-instance-template  \
     --source-instance=my-source-instance \
     --configure-disk=device-name=data-disk-a,instantiate-from=source-image,auto-delete=true \
     --configure-disk=device-name=data-disk-b,instantiate-from=custom-image,custom-image=projects/cps-cloud/global/images/cos-89-16108-403-15

Go 

import (
	"context"
	"fmt"
	"io"

	compute "cloud.google.com/go/compute/apiv1"
	computepb "cloud.google.com/go/compute/apiv1/computepb"
	"google.golang.org/protobuf/proto"
)

// createTemplateFromInstance creates a new instance template based on an existing instance.
// This new template specifies a different boot disk.
func createTemplateFromInstance(w io.Writer, projectID, instance, templateName string) error {
	// projectID := "your_project_id"
	// instance := "projects/project/zones/zone/instances/instance"
	// templateName := "your_template_name"

	ctx := context.Background()
	instanceTemplatesClient, err := compute.NewInstanceTemplatesRESTClient(ctx)
	if err != nil {
		return fmt.Errorf("NewInstanceTemplatesRESTClient: %w", err)
	}
	defer instanceTemplatesClient.Close()

	req := &computepb.InsertInstanceTemplateRequest{
		Project: projectID,
		InstanceTemplateResource: &computepb.InstanceTemplate{
			Name:           proto.String(templateName),
			SourceInstance: proto.String(instance),
			SourceInstanceParams: &computepb.SourceInstanceParams{
				DiskConfigs: []*computepb.DiskInstantiationConfig{
					{
						// Device name must match the name of a disk attached to the instance
						// your template is based on.
						DeviceName: proto.String("disk-1"),
						// Replace the original boot disk image used in your instance with a Rocky Linux image.
						InstantiateFrom: proto.String(computepb.DiskInstantiationConfig_CUSTOM_IMAGE.String()),
						CustomImage:     proto.String("projects/rocky-linux-cloud/global/images/family/rocky-linux-8"),
						// Override the auto_delete setting.
						AutoDelete: proto.Bool(true),
					},
				},
			},
		},
	}

	op, err := instanceTemplatesClient.Insert(ctx, req)
	if err != nil {
		return fmt.Errorf("unable to create instance template: %w", err)
	}

	if err = op.Wait(ctx); err != nil {
		return fmt.Errorf("unable to wait for the operation: %w", err)
	}

	fmt.Fprintf(w, "Instance template created\n")

	return nil
}

Java 


import com.google.cloud.compute.v1.DiskInstantiationConfig;
import com.google.cloud.compute.v1.DiskInstantiationConfig.InstantiateFrom;
import com.google.cloud.compute.v1.GlobalOperationsClient;
import com.google.cloud.compute.v1.InsertInstanceTemplateRequest;
import com.google.cloud.compute.v1.InstanceTemplate;
import com.google.cloud.compute.v1.InstanceTemplatesClient;
import com.google.cloud.compute.v1.Operation;
import com.google.cloud.compute.v1.SourceInstanceParams;
import java.io.IOException;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;

public class CreateTemplateFromInstance {

  public static void main(String[] args)
      throws IOException, ExecutionException, InterruptedException, TimeoutException {
    // TODO(developer): Replace these variables before running the sample.
    // projectId: project ID or project number of the Cloud project you use.
    // instance: the instance to base the new template on. This value uses the following format:
    // **NOTE**: "projects/{project}/zones/{zone}/instances/{instance_name}"
    // templateName: name of the new template to create.
    String projectId = "your-project-id";
    String templateName = "template-name";
    String instance = String.format("projects/%s/zones/%s/instances/%s", projectId, "zone",
        "instanceName");
    createTemplateFromInstance(projectId, templateName, instance);
  }

  // Create a new instance template based on an existing instance.
  // This new template specifies a different boot disk.
  public static void createTemplateFromInstance(String projectId, String templateName,
      String instance)
      throws IOException, ExecutionException, InterruptedException, TimeoutException {
    try (InstanceTemplatesClient instanceTemplatesClient = InstanceTemplatesClient.create();
        GlobalOperationsClient globalOperationsClient = GlobalOperationsClient.create()) {

      SourceInstanceParams sourceInstanceParams = SourceInstanceParams.newBuilder()
          .addDiskConfigs(DiskInstantiationConfig.newBuilder()
              // Device name must match the name of a disk attached to the instance you are
              // basing your template on.
              .setDeviceName("disk-1")
              // Replace the original boot disk image used in your instance
              // with a Rocky Linux image.
              .setInstantiateFrom(InstantiateFrom.CUSTOM_IMAGE.toString())
              .setCustomImage(
                  String.format("projects/%s/global/images/family/%s", "rocky-linux-cloud",
                      "rocky-linux-8"))
              // Override the AutoDelete setting.
              .setAutoDelete(true).build())
          .build();

      InstanceTemplate instanceTemplate = InstanceTemplate.newBuilder()
          .setName(templateName)
          .setSourceInstance(instance)
          .setSourceInstanceParams(sourceInstanceParams)
          .build();

      InsertInstanceTemplateRequest insertInstanceTemplateRequest = InsertInstanceTemplateRequest
          .newBuilder()
          .setProject(projectId)
          .setInstanceTemplateResource(instanceTemplate)
          .build();

      Operation operation = instanceTemplatesClient.insertCallable()
          .futureCall(insertInstanceTemplateRequest).get(3, TimeUnit.MINUTES);

      Operation response = globalOperationsClient.wait(projectId, operation.getName());

      if (response.hasError()) {
        System.out.println("Instance Template creation failed ! ! " + response);
        return;
      }
      System.out.printf("Instance Template creation operation status %s: %s", templateName,
          response.getStatus());
    }
  }
}

Node.js 

/**
 * TODO(developer): Uncomment and replace these variables before running the sample.
 */
// const projectId = 'YOUR_PROJECT_ID';
// const instance = 'projects/project/zones/zone/instances/instance';
// const templateName = 'your_template_name';

const compute = require('@google-cloud/compute');

// Create a new instance template based on an existing instance.
// This new template specifies a different boot disk.
async function createTemplateFromInstance() {
  const instanceTemplatesClient = new compute.InstanceTemplatesClient();

  const [response] = await instanceTemplatesClient.insert({
    project: projectId,
    instanceTemplateResource: {
      name: templateName,
      sourceInstance: instance,
      sourceInstanceParams: {
        diskConfigs: [
          {
            // Device name must match the name of a disk attached to the instance
            // your template is based on.
            deviceName: 'disk-1',
            // Replace the original boot disk image used in your instance with a Rocky Linux image.
            instantiateFrom: 'CUSTOM_IMAGE',
            customImage:
              'projects/rocky-linux-cloud/global/images/family/rocky-linux-8',
            // Override the auto_delete setting.
            autoDelete: true,
          },
        ],
      },
    },
  });
  let operation = response.latestResponse;
  const operationsClient = new compute.GlobalOperationsClient();

  // Wait for the create operation to complete.
  while (operation.status !== 'DONE') {
    [operation] = await operationsClient.wait({
      operation: operation.name,
      project: projectId,
    });
  }

  console.log('Instance template created.');
}

createTemplateFromInstance();

Python 

from __future__ import annotations

import sys
from typing import Any

from google.api_core.extended_operation import ExtendedOperation
from google.cloud import compute_v1


def wait_for_extended_operation(
    operation: ExtendedOperation, verbose_name: str = "operation", timeout: int = 300
) -> Any:
    """
    Waits for the extended (long-running) operation to complete.

    If the operation is successful, it will return its result.
    If the operation ends with an error, an exception will be raised.
    If there were any warnings during the execution of the operation
    they will be printed to sys.stderr.

    Args:
        operation: a long-running operation you want to wait on.
        verbose_name: (optional) a more verbose name of the operation,
            used only during error and warning reporting.
        timeout: how long (in seconds) to wait for operation to finish.
            If None, wait indefinitely.

    Returns:
        Whatever the operation.result() returns.

    Raises:
        This method will raise the exception received from `operation.exception()`
        or RuntimeError if there is no exception set, but there is an `error_code`
        set for the `operation`.

        In case of an operation taking longer than `timeout` seconds to complete,
        a `concurrent.futures.TimeoutError` will be raised.
    """
    result = operation.result(timeout=timeout)

    if operation.error_code:
        print(
            f"Error during {verbose_name}: [Code: {operation.error_code}]: {operation.error_message}",
            file=sys.stderr,
            flush=True,
        )
        print(f"Operation ID: {operation.name}", file=sys.stderr, flush=True)
        raise operation.exception() or RuntimeError(operation.error_message)

    if operation.warnings:
        print(f"Warnings during {verbose_name}:\n", file=sys.stderr, flush=True)
        for warning in operation.warnings:
            print(f" - {warning.code}: {warning.message}", file=sys.stderr, flush=True)

    return result


def create_template_from_instance(
    project_id: str, instance: str, template_name: str
) -> compute_v1.InstanceTemplate:
    """
    Create a new instance template based on an existing instance.
    This new template specifies a different boot disk.

    Args:
        project_id: project ID or project number of the Cloud project you use.
        instance: the instance to base the new template on. This value uses
            the following format: "projects/{project}/zones/{zone}/instances/{instance_name}"
        template_name: name of the new template to create.

    Returns:
        InstanceTemplate object that represents the new instance template.
    """
    disk = compute_v1.DiskInstantiationConfig()
    # Device name must match the name of a disk attached to the instance you are
    # basing your template on.
    disk.device_name = "disk-1"
    # Replace the original boot disk image used in your instance with a Rocky Linux image.
    disk.instantiate_from = "CUSTOM_IMAGE"
    disk.custom_image = "projects/rocky-linux-cloud/global/images/family/rocky-linux-8"
    # Override the auto_delete setting.
    disk.auto_delete = True

    template = compute_v1.InstanceTemplate()
    template.name = template_name
    template.source_instance = instance
    template.source_instance_params = compute_v1.SourceInstanceParams()
    template.source_instance_params.disk_configs = [disk]

    template_client = compute_v1.InstanceTemplatesClient()
    operation = template_client.insert(
        project=project_id, instance_template_resource=template
    )

    wait_for_extended_operation(operation, "instance template creation")

    return template_client.get(project=project_id, instance_template=template_name)

REST

To create a regional instance template, use the regionInstanceTemplates.insert method, or, to create a global instance template, use the instanceTemplates.insert method.

In your request, you must specify the sourceInstance field. To override how the source instance's disks are defined, add one or more diskConfigs fields.

For example, make the following call to create a global instance template from an existing instance.

POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/global/instanceTemplates

{
  "name": "INSTANCE_TEMPLATE_NAME",
  "sourceInstance": "zones/SOURCE_INSTANCE_ZONE/instances/SOURCE_INSTANCE",
  "sourceInstanceParams": {
    "diskConfigs": [
      {
        "deviceName": "SOURCE_DISK",
        "instantiateFrom": "INSTANTIATE_OPTIONS",
        "autoDelete": false
      }
    ]
  }
}

Replace the following:

  • PROJECT_ID: your project id
  • INSTANCE_TEMPLATE_NAME: the name for the new template
  • SOURCE_INSTANCE_ZONE: the zone of the source instance
  • SOURCE_INSTANCE: the name of the source instance to use as a model for this instance template
  • SOURCE_DISK: the name of a source-instance disk that you want to override within the template

  • INSTANTIATE_OPTIONS: specifies whether to include the disk and which image to use

    Valid values depend on the type of disk:

    • source-image or source-image-family (valid only for boot and other persistent read-write disks).
    • custom-image (valid only for boot and other persistent read-write disks). If you want to retain applications and settings from the source VMs in your instance template, you can create a custom image and then specify it when you create the template. If specified, then provide the path or URL for the custom image, as shown in the following example. Alternatively, you can specify an image family using the following format:
      --configure-disk=device-name=DATA_DISK_NAME,instantiate-from=custom-image, \
  custom-image=projects/PROJECT_ID/global/images/family/IMAGE_FAMILY_NAME
  ```

+ `attach-read-only` (valid only for read-only disks).
+ `blank` (valid only for non-boot persistent disks and local SSDs). If
specified, then, when the template is used to create a new instance, the
disk is created unformatted. You must
[format and mount](/compute/docs/disks/format-mount-disk-linux#format_linux)
the disk in a startup script before you can use it in a scalable setup.
+ `do-not-include` (valid only for non-boot persistent disks and
read-only disks).

The following example creates a new instance template based on my-source-instance. In the instance template, the image for data-disk-a is replaced with projects/cos-cloud/global/images/cos-89-16108-403-15.

POST https://compute.googleapis.com/compute/v1/projects/my_project/global/instanceTemplates

{
  "name": "my-instance-template",
  "sourceInstance": "zones/us-central1-a/instances/my-source-instance",
  "sourceInstanceParams":
  {
    "diskConfigs":
    [
      {
        "deviceName": "data-disk-a",
        "instantiateFrom": "custom-image",
        "customImage": "projects/cos-cloud/global/images/cos-89-16108-403-15"
      }
    ]
  }
}

The following table shows how the options for overriding disks are defined in the template.

Disk type Options
Boot disk
  • [Default] Use the same source image or image family that was used to create the boot disk in the source instance.
  • Use the URL of any image (custom or public) as described in the preceding example or specify an image family using the following format:

    • projects/exampleproject/global/images/family/IMAGE_FAMILY_NAME

Other read/write persistent disks
  • [Default] Use the same source image/source image family that was used to create the disk in the source instance. Note: If the source instance's disk doesn't have a source image/source image family property, then it is included in the template as a blank disk.
  • Use the URL of any image (custom or public) as described in the preceding example or specify an image family using the following format:

    projects/exampleproject/global/images/family/IMAGE_FAMILY_NAME

  • Use a blank disk in the template instead. When the template is used to create a new instance, this disk is created unformatted. You must format and mount the disk in a startup script before you can use it in a scalable setup.
  • Don't include the disk.
Read-only disk(s)
  • [Default] Include the disk in read-only mode.
  • Don't include the disk.
Local SSD(s)
  • [Default] Include a blank local SSD. When the template is used to create a new instance, this disk is created unformatted. You must format and mount the disk in a startup script before you can use it in a scalable setup.

For each disk, you can also override the auto-delete attribute to specify whether the disk should be deleted when its associated instance is deleted.

By default, if no override options are specified, the disk configuration in the template matches the source instance.

Create an instance template based on an existing template

You can't update an existing instance template. But, if an instance template goes out of date or if you need to make changes, you can create another one with similar properties by using the console.

  1. Go to the Instance templates page.

    Go to Instance templates

  2. Click the instance template that you want to copy and update.

  3. Click Create similar.

  4. Update the configuration in the new template.

  5. Click Create.

Create an instance template for GPU VMs

When creating an instance template, you can configure it for creating VMs that have attached GPUs by specifying the following:

Console

To create an instance template for GPU VMs, do the following:

  1. In the Google Cloud console, go to the Instance templates page.

    Go to Instance templates

  2. Click Create instance template.

  3. In the Name field, enter a name for the instance template.

  4. In the Location section, select one of the following options:

    • To create a global instance template, select Global (default).

    • To create a regional instance template, select Regional, and then select the region where you want to create the instance template.

  5. In the Machine configuration section, do the following:

    1. Click the GPUs tab.

    2. In the GPU type menu, select the GPU type.

    3. In the Number of GPUs menu, select the number of GPUs.

    4. Optional: If your GPU model supports NVIDIA RTX Virtual Workstations (vWS) for graphics workloads, and you plan to run graphics-intensive workloads, select Enable Virtual Workstation (NVIDIA GRID).

    5. In the Machine type section, select a machine type.

  6. Optional: To change the default value boot disk type or image, in the Boot disk section, click Change. Then, follow the prompts to change the boot disk.

  7. Click Create.

gcloud

To create an instance template for GPU VMs, use the instance-templates create command with the --maintenance-policy flag set to TERMINATE.

For example, to create a global instance template for GPU VMs, use the following command:

gcloud compute instance-templates create INSTANCE_TEMPLATE_NAME \
    --image-project=IMAGE_PROJECT \
    --image-family=IMAGE_FAMILY \
    --machine-type=MACHINE_TYPE \
    --maintenance-policy=TERMINATE

Replace the following:

  • INSTANCE_TEMPLATE_NAME: the name of the instance template.

  • IMAGE_PROJECT: the image project that contains the image; for example, debian-cloud. For more information about the supported image projects, see Public images.

  • IMAGE_FAMILY or IMAGE: specify one of the following:

    • IMAGE_FAMILY: an image family. This specifies the most recent, non-deprecated OS image. For example, if you specify debian-10, the latest version in the Debian 10 image family is used. For more information about using image families, see Image families best practices.

    • IMAGE: a specific version of the OS image; for example, debian-10-buster-v20200309. If you choose to specify a specific version of the OS image, then you must replace the --image-family flag with the --image flag.

  • MACHINE_TYPE: the machine type of the VMs. If you specify an N1 machine type, then include the --accelerator flag to specify the number and type of GPUs to attach to your VMs.

For example, assume that you want to create a global instance template for GPU VMs that specifies the following properties:

  • An N1 predefined machine type with 2 vCPUs.

  • One NVIDIA T4 GPU to attach to the VMs.

  • Debian as the image project.

  • Debian 10 as the image family.

To create the example instance template, use the following command:

gcloud compute instance-templates create instance-template-gpu \
    --accelerator=count=1,type=nvidia-tesla-t4 \
    --machine-type=n1-standard-2 \
    --image-family=debian-10 \
    --image-project=debian-cloud \
    --maintenance-policy=TERMINATE

Terraform

To create an instance template for GPU VMs, use the google_compute_region_instance_template resource.

For example, to create a global instance template, which specifies an N1 predefined machine type with 2 vCPUs and one NVIDIA T4 GPU attached, use the following resource:

resource "google_compute_instance_template" "default" {
  name         = "gpu-template"
  machine_type = "n1-standard-2"

  disk {
    source_image = "debian-cloud/debian-11"
  }

  network_interface {
    network = "default"
  }

  guest_accelerator {
    type  = "nvidia-tesla-t4"
    count = 1
  }

  scheduling {
    on_host_maintenance = "TERMINATE"
  }
}

To learn how to apply or remove a Terraform configuration, see Basic Terraform commands.

REST

To create an instance template for GPU VMs, make a POST request to the instanceTemplates.insert method. In the request body, include the onHostMaintenance field and set it to TERMINATE.

For example, to create a global instance template for GPU VMs, make a POST request as follows:

POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/global/instanceTemplates

{
  "name": "INSTANCE_TEMPLATE_NAME",
  "properties": {
    "disks": [
      {
        "type": "PERSISTENT",
        "boot": true,
        "mode": "READ_WRITE",
        "initializeParams": {
          "sourceImage": "projects/IMAGE_PROJECT/global/images/IMAGE"
        }
      }
    ],
    "machineType": "MACHINE_TYPE",
    "networkInterfaces": [
      {
        "accessConfigs": [
          {
            "name": "external-IP",
            "type": "ONE_TO_ONE_NAT"
          }
        ],
        "network": "global/networks/default"
      }
    ],
    "scheduling": {
      "onHostMaintenance": "TERMINATE"
    }
  }
}

Replace the following:

  • PROJECT_ID: the ID of the project in which you want to create the instance template.

  • INSTANCE_TEMPLATE_NAME: the name of the instance template.

  • IMAGE_PROJECT: the image project that contains the image; for example, debian-cloud. For more information about the supported image projects, see Public images.

  • IMAGE or IMAGE_FAMILY: specify one of the following:

    • IMAGE: a specific version of the OS image; for example, debian-10-buster-v20200309.

    • IMAGE_FAMILY: an image family. This specifies the most recent, non-deprecated OS image. For example, if you specify family/debian-10, the latest version in the Debian 10 image family is used. For more information about using image families, see Image families best practices.

  • MACHINE_TYPE: the machine type of the VMs. If you specify an N1 machine type, then include the guestAccelerators field to specify the number and type of GPUs to attach to your VMs.

For example, assume that you want to create a global instance template for GPU VMs that specifies the following properties:

  • An N1 predefined machine type with 2 vCPUs.

  • One NVIDIA T4 GPU to attach to the VMs.

  • Debian as the image project.

  • Debian 10 as the image family.

To create the example instance template, make a POST request as follows:

POST https://compute.googleapis.com/compute/v1/projects/example-project/global/instanceTemplates

{
  "name": "instance-template-gpu",
  "properties": {
    "disks": [
      {
        "type": "PERSISTENT",
        "boot": true,
        "mode": "READ_WRITE",
        "initializeParams": {
          "sourceImage": "projects/debian-cloud/global/images/family/debian-10"
        }
      }
    ],
    "guestAccelerators": [
      {
        "acceleratorType": "nvidia-tesla-t4",
        "acceleratorCount": 1
      }
    ],
    "machineType": "n1-standard-2",
    "networkInterfaces": [
      {
        "accessConfigs": [
          {
            "name": "external-IP",
            "type": "ONE_TO_ONE_NAT"
          }
        ],
        "network": "global/networks/default"
      }
    ],
    "scheduling": {
      "onHostMaintenance": "TERMINATE"
    }
  }
}

For more configuration options when creating an instance template, see Create an instance template in this document.

Create an instance template with a container image

You can specify a container image in an instance template. By default, Compute Engine also includes in the template a Container-Optimized OS image with Docker installed. When you use the template to create a new instance, the container is launched automatically as the instance starts up.

Console

  1. Go to the Instance templates page.

    Go to Instance templates

  2. Click Create instance template.

  3. In the Container section, click Deploy Container.

  4. In the Configure container dialog, specify the Container image to use.

    • You can specify an image from Container Registry or Artifact Registry. For example:
      • gcr.io/cloud-marketplace/google/nginx1:TAG, where TAG is the tag defined for a specific version of NGINX container image available on Google Cloud Marketplace.
      • us-docker.pkg.dev/google-samples/containers/gke/hello-app:1.0 selects a sample hello-app image stored in Artifact Registry.
    • If you use a container image from Docker Hub, always specify the full Docker image name. For example, specify the following image name to deploy an Apache container image: docker.io/httpd:2.4.

  5. Optionally, click Advanced container options. For more information, see Configuring options to run your Container.

  6. Click Create.

gcloud

Use the gcloud compute instance-templates create-with-container command:

 gcloud compute instance-templates create-with-container INSTANCE_TEMPLATE_NAME \
     --container-image=CONTAINER_IMAGE

Replace the following:

  • INSTANCE_TEMPLATE_NAME: The name of the template to create.
  • CONTAINER_IMAGE: The full name of the container image to use.

For example, the following command creates a new instance template named nginx-vm. A VM instance created from this template launches and runs the container image, gcr.io/cloud-marketplace/google/nginx1:TAG, when the VM starts.

 gcloud compute instance-templates create-with-container nginx-vm \
     --container-image=gcr.io/cloud-marketplace/google/nginx1:TAG

Replace TAG with the tag defined for a specific version of NGINX container image available on Google Cloud Marketplace.

You can also configure options to run your container.

Create an instance template that specifies a subnet

gcloud

To create a regional or global instance template, use the instance-templates create command. Use the --subnet flag to place instances that are created from the template into the subnet of your choice. The --subnet flag requires the --region flag.

If you want to create a regional instance template, you must use the --instance-template-region flag to set the region of the template. Make sure that you use a subnet from the same region as where you want to create the regional instance template.

gcloud compute instance-templates create INSTANCE_TEMPLATE_NAME \
    --region=REGION \
    --subnet=SUBNET_NAME_OR_URL \
    --stack-type=STACK_TYPE \
    --instance-template-region=INSTANCE_TEMPLATE_REGION

Replace the following:

  • INSTANCE_TEMPLATE_NAME: the name for the instance template
  • REGION: the region of the subnet

  • SUBNET_NAME_OR_URL: either the name of the subnet or its URL

  • STACK_TYPE: Optional: whether IPv6 is enabled on the default network interface. The following values can be used: IPV4_ONLY, IPV4_IPV6, or IPV6_ONLY (Preview). If you don't include this flag, the default value is IPV4_ONLY.

  • INSTANCE_TEMPLATE_REGION: the region where you want to create the instance template. This region must be same as REGION.

The following example creates a template called template-qa that only creates instances in the subnet-us-qa subnet.

gcloud compute instance-templates create template-qa \
    --region=us-central1 \
    --subnet=subnet-us-qa

The output is similar to the following:

Created [https://compute.googleapis.com/compute/latest/projects/PROJECT_ID/global/instanceTemplates/template-qa].
NAME        MACHINE_TYPE        PREEMPTIBLE CREATION_TIMESTAMP
template-qa e2-standard-2       2019-12-23T20:34:00.791-07:00

Go 

import (
	"context"
	"fmt"
	"io"

	compute "cloud.google.com/go/compute/apiv1"
	computepb "cloud.google.com/go/compute/apiv1/computepb"
	"google.golang.org/protobuf/proto"
)

// createTemplateWithSubnet creates an instance template that uses a provided subnet.
func createTemplateWithSubnet(w io.Writer, projectID, network, subnetwork, templateName string) error {
	// projectID := "your_project_id"
	// network := "projects/project/global/networks/network"
	// subnetwork := "projects/project/regions/region/subnetworks/subnetwork"
	// templateName := "your_template_name"

	ctx := context.Background()
	instanceTemplatesClient, err := compute.NewInstanceTemplatesRESTClient(ctx)
	if err != nil {
		return fmt.Errorf("NewInstanceTemplatesRESTClient: %w", err)
	}
	defer instanceTemplatesClient.Close()

	req := &computepb.InsertInstanceTemplateRequest{
		Project: projectID,
		InstanceTemplateResource: &computepb.InstanceTemplate{
			Name: proto.String(templateName),
			Properties: &computepb.InstanceProperties{
				// The template describes the size and source image of the boot disk
				// to attach to the instance.
				Disks: []*computepb.AttachedDisk{
					{
						InitializeParams: &computepb.AttachedDiskInitializeParams{
							DiskSizeGb:  proto.Int64(250),
							SourceImage: proto.String("projects/debian-cloud/global/images/family/debian-11"),
						},
						AutoDelete: proto.Bool(true),
						Boot:       proto.Bool(true),
					},
				},
				MachineType: proto.String("e2-standard-4"),
				// The template connects the instance to the specified network and subnetwork.
				NetworkInterfaces: []*computepb.NetworkInterface{
					{
						Network:    proto.String(network),
						Subnetwork: proto.String(subnetwork),
					},
				},
			},
		},
	}

	op, err := instanceTemplatesClient.Insert(ctx, req)
	if err != nil {
		return fmt.Errorf("unable to create instance template: %w", err)
	}

	if err = op.Wait(ctx); err != nil {
		return fmt.Errorf("unable to wait for the operation: %w", err)
	}

	fmt.Fprintf(w, "Instance template created\n")

	return nil
}

Java 


import com.google.cloud.compute.v1.AttachedDisk;
import com.google.cloud.compute.v1.AttachedDiskInitializeParams;
import com.google.cloud.compute.v1.GlobalOperationsClient;
import com.google.cloud.compute.v1.InsertInstanceTemplateRequest;
import com.google.cloud.compute.v1.InstanceProperties;
import com.google.cloud.compute.v1.InstanceTemplate;
import com.google.cloud.compute.v1.InstanceTemplatesClient;
import com.google.cloud.compute.v1.NetworkInterface;
import com.google.cloud.compute.v1.Operation;
import java.io.IOException;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;

public class CreateTemplateWithSubnet {

  public static void main(String[] args)
      throws IOException, ExecutionException, InterruptedException, TimeoutException {
    /*
    TODO(developer): Replace these variables before running the sample.
    projectId: project ID or project number of the Cloud project you use.
    network: the network to be used in the new template. This value uses
        the following format: "projects/{project}/global/networks/{network}"
    subnetwork: the subnetwork to be used in the new template. This value
        uses the following format: "projects/{project}/regions/{region}/subnetworks/{subnetwork}"
    templateName: name of the new template to create.
    */
    String projectId = "your-project-id";
    String network = String.format("projects/%s/global/networks/%s", projectId, "network");
    String subnetwork = String.format("projects/%s/regions/%s/subnetworks/%s", projectId, "region",
        "subnetwork");
    String templateName = "template-name";
    createTemplateWithSubnet(projectId, network, subnetwork, templateName);
  }

  // Create an instance template that uses a provided subnet.
  public static void createTemplateWithSubnet(String projectId, String network, String subnetwork,
      String templateName)
      throws IOException, ExecutionException, InterruptedException, TimeoutException {
    try (InstanceTemplatesClient instanceTemplatesClient = InstanceTemplatesClient.create();
        GlobalOperationsClient globalOperationsClient = GlobalOperationsClient.create()) {

      AttachedDisk disk = AttachedDisk.newBuilder()
          .setInitializeParams(AttachedDiskInitializeParams.newBuilder()
              .setSourceImage(
                  String.format("projects/%s/global/images/family/%s", "debian-cloud", "debian-11"))
              .setDiskSizeGb(250).build())
          .setAutoDelete(true)
          .setBoot(true)
          .build();

      InstanceProperties instanceProperties = InstanceProperties.newBuilder()
          .addDisks(disk)
          .setMachineType("e2-standard-4")
          .addNetworkInterfaces(NetworkInterface.newBuilder()
              .setNetwork(network)
              .setSubnetwork(subnetwork).build())
          .build();

      InstanceTemplate instanceTemplate = InstanceTemplate.newBuilder()
          .setName(templateName)
          .setProperties(instanceProperties)
          .build();

      InsertInstanceTemplateRequest insertInstanceTemplateRequest = InsertInstanceTemplateRequest
          .newBuilder()
          .setProject(projectId)
          .setInstanceTemplateResource(instanceTemplate)
          .build();

      Operation operation = instanceTemplatesClient.insertCallable()
          .futureCall(insertInstanceTemplateRequest).get(3, TimeUnit.MINUTES);

      Operation response = globalOperationsClient.wait(projectId, operation.getName());

      if (response.hasError()) {
        System.out.println("Template creation from subnet failed ! ! " + response);
        return;
      }
      System.out.printf("Template creation from subnet operation status %s: %s", templateName,
          response.getStatus());
    }
  }
}

Node.js 

/**
 * TODO(developer): Uncomment and replace these variables before running the sample.
 */
// const projectId = 'YOUR_PROJECT_ID';
// const network = 'projects/project/global/networks/network';
// const subnetwork = 'projects/project/regions/region/subnetworks/subnetwork';
// const templateName = 'your_template_name';

const compute = require('@google-cloud/compute');

// Create an instance template that uses a provided subnet.
async function createTemplateWithSubnet() {
  const instanceTemplatesClient = new compute.InstanceTemplatesClient();

  const [response] = await instanceTemplatesClient.insert({
    project: projectId,
    instanceTemplateResource: {
      name: templateName,
      properties: {
        // The template describes the size and source image of the boot disk
        // to attach to the instance.
        disks: [
          {
            // The template describes the size and source image of the boot disk
            // to attach to the instance.
            initializeParams: {
              diskSizeGb: '250',
              sourceImage:
                'projects/debian-cloud/global/images/family/debian-11',
            },
            autoDelete: true,
            boot: true,
          },
        ],
        machineType: 'e2-standard-4',
        // The template connects the instance to the specified network and subnetwork.
        networkInterfaces: [
          {
            network,
            subnetwork,
          },
        ],
      },
    },
  });
  let operation = response.latestResponse;
  const operationsClient = new compute.GlobalOperationsClient();

  // Wait for the create operation to complete.
  while (operation.status !== 'DONE') {
    [operation] = await operationsClient.wait({
      operation: operation.name,
      project: projectId,
    });
  }

  console.log('Instance template created.');
}

createTemplateWithSubnet();

Python 

from __future__ import annotations

import sys
from typing import Any

from google.api_core.extended_operation import ExtendedOperation
from google.cloud import compute_v1


def wait_for_extended_operation(
    operation: ExtendedOperation, verbose_name: str = "operation", timeout: int = 300
) -> Any:
    """
    Waits for the extended (long-running) operation to complete.

    If the operation is successful, it will return its result.
    If the operation ends with an error, an exception will be raised.
    If there were any warnings during the execution of the operation
    they will be printed to sys.stderr.

    Args:
        operation: a long-running operation you want to wait on.
        verbose_name: (optional) a more verbose name of the operation,
            used only during error and warning reporting.
        timeout: how long (in seconds) to wait for operation to finish.
            If None, wait indefinitely.

    Returns:
        Whatever the operation.result() returns.

    Raises:
        This method will raise the exception received from `operation.exception()`
        or RuntimeError if there is no exception set, but there is an `error_code`
        set for the `operation`.

        In case of an operation taking longer than `timeout` seconds to complete,
        a `concurrent.futures.TimeoutError` will be raised.
    """
    result = operation.result(timeout=timeout)

    if operation.error_code:
        print(
            f"Error during {verbose_name}: [Code: {operation.error_code}]: {operation.error_message}",
            file=sys.stderr,
            flush=True,
        )
        print(f"Operation ID: {operation.name}", file=sys.stderr, flush=True)
        raise operation.exception() or RuntimeError(operation.error_message)

    if operation.warnings:
        print(f"Warnings during {verbose_name}:\n", file=sys.stderr, flush=True)
        for warning in operation.warnings:
            print(f" - {warning.code}: {warning.message}", file=sys.stderr, flush=True)

    return result


def create_template_with_subnet(
    project_id: str, network: str, subnetwork: str, template_name: str
) -> compute_v1.InstanceTemplate:
    """
    Create an instance template that uses a provided subnet.

    Args:
        project_id: project ID or project number of the Cloud project you use.
        network: the network to be used in the new template. This value uses
            the following format: "projects/{project}/global/networks/{network}"
        subnetwork: the subnetwork to be used in the new template. This value
            uses the following format: "projects/{project}/regions/{region}/subnetworks/{subnetwork}"
        template_name: name of the new template to create.

    Returns:
        InstanceTemplate object that represents the new instance template.
    """
    # The template describes the size and source image of the book disk to
    # attach to the instance.
    disk = compute_v1.AttachedDisk()
    initialize_params = compute_v1.AttachedDiskInitializeParams()
    initialize_params.source_image = (
        "projects/debian-cloud/global/images/family/debian-11"
    )
    initialize_params.disk_size_gb = 250
    disk.initialize_params = initialize_params
    disk.auto_delete = True
    disk.boot = True

    template = compute_v1.InstanceTemplate()
    template.name = template_name
    template.properties = compute_v1.InstanceProperties()
    template.properties.disks = [disk]
    template.properties.machine_type = "e2-standard-4"

    # The template connects the instance to the specified network and subnetwork.
    network_interface = compute_v1.NetworkInterface()
    network_interface.network = network
    network_interface.subnetwork = subnetwork
    template.properties.network_interfaces = [network_interface]

    template_client = compute_v1.InstanceTemplatesClient()
    operation = template_client.insert(
        project=project_id, instance_template_resource=template
    )
    wait_for_extended_operation(operation, "instance template creation")

    return template_client.get(project=project_id, instance_template=template_name)

Using this template to create instances for a MIG (with or without autoscaling) automatically creates the instance in the specified region and subnet. This lets you control the subnet of new instances created for load balancing.

Use custom or public images in your instance templates

You can either use a custom image or a public image for your instance templates:

  • Custom images. As MIGs are designed to add and remove instances frequently, it is useful to create a custom image and specify it in the instance template. You can prepare your image with the applications and settings that your VMs need, so you don't have to manually configure those items on individual VMs in the MIG.

  • Public images. You can create an instance template that uses a public image and a startup script to prepare the instance after it starts running.

Custom images are more deterministic and start more quickly than VMs with startup scripts. However, startup scripts are more flexible, which helps you update the apps and settings in your instances.

If you're managing images using image families, you can specify the name of your custom or public image family in the instance template. For more information on image families, see best practices when using image families on Compute Engine.

What's next