Define job resources using a VM instance template

This document explains how to define the VM resources for a Batch job by specifying a Compute Engine VM instance template when you create the job.

The types of VM resources that a job runs on are automatically defined by Batch unless you define them by using one of the following methods:

  • Define a job's VM resources directly by using the instances[].policy field. This method is demonstrated in most Batch documentation.
  • Define a job's VM resources through a template by using the instances[].instanceTemplate field. This is the method explained in this document.

    Using a template is required to create a job that uses non-default VM images. Using a template can also be convenient when you want to specify the same VM resources for multiple jobs.

Before you begin

  1. If you haven't used Batch before, review Get started with Batch and enable Batch by completing the prerequisites for projects and users.
  2. Create an instance template or identify an existing instance template.
  3. To get the permissions that you need to create a job, ask your administrator to grant you the following IAM roles:

    For more information about granting roles, see Manage access to projects, folders, and organizations.

    You might also be able to get the required permissions through custom roles or other predefined roles.

Create a job using a Compute Engine VM instance template

This section provides examples for how to create a basic script job from an existing VM instance template. You can create a job from a VM instance template using the gcloud CLI, Batch API, Go, Java, Node.js, Python, or C++.

gcloud

To create a job from a VM instance template using the gcloud CLI, use the gcloud batch jobs submit command and specify the VM instance template in the job's JSON configuration file.

For example, to create a basic script job from a VM instance template:

  1. Create a JSON file in the current directory named hello-world-instance-template.json with the following contents:

    {
        "taskGroups": [
            {
                "taskSpec": {
                    "runnables": [
                        {
                            "script": {
                                "text": "echo Hello world! This is task ${BATCH_TASK_INDEX}. This job has a total of ${BATCH_TASK_COUNT} tasks."
                            }
                        }
                    ],
                    "computeResource": {
                        "cpuMilli": 2000,
                        "memoryMib": 16
                    },
                    "maxRetryCount": 2,
                    "maxRunDuration": "3600s"
                },
                "taskCount": 4,
                "parallelism": 2
            }
        ],
        "allocationPolicy": {
            "instances": [
                {
                    "installGpuDrivers": INSTALL_GPU_DRIVERS,
                    "instanceTemplate": "INSTANCE_TEMPLATE_NAME"
                }
            ]
        },
        "labels": {
            "department": "finance",
            "env": "testing"
        },
        "logsPolicy": {
            "destination": "CLOUD_LOGGING"
        }
    }
    

    Replace the following:

    • INSTALL_GPU_DRIVERS: Optional. When set to true, Batch fetches the drivers required for the GPU type that you specify in your Compute Engine VM instance template, and Batch installs them on your behalf. For more information, see how to create a job that uses a GPU.
    • INSTANCE_TEMPLATE_NAME: the name of an existing Compute Engine VM instance template. Learn how to create and list instance templates.
  2. Run the following command:

    gcloud batch jobs submit example-template-job \
      --location us-central1 \
      --config hello-world-instance-template.json
    

API

To create a basic job using the Batch API, use the jobs.create method and specify a VM instance template in the allocationPolicy field.

For example, to create a basic script jobs from a VM instance template, use the following request:

POST https://batch.googleapis.com/v1/projects/PROJECT_ID/locations/us-central1/jobs?job_id=example-script-job

{
    "taskGroups": [
        {
            "taskSpec": {
                "runnables": [
                    {
                        "script": {
                            "text": "echo Hello world! This is task ${BATCH_TASK_INDEX}. This job has a total of ${BATCH_TASK_COUNT} tasks."
                        }
                    }
                ],
                "computeResource": {
                    "cpuMilli": 2000,
                    "memoryMib": 16
                },
                "maxRetryCount": 2,
                "maxRunDuration": "3600s"
            },
            "taskCount": 4,
            "parallelism": 2
        }
    ],
    "allocationPolicy": {
        "instances": [
            {
                "installGpuDrivers": INSTALL_GPU_DRIVERS,
                "instanceTemplate": "INSTANCE_TEMPLATE_NAME"
            }
        ]
    },
    "labels": {
        "department": "finance",
        "env": "testing"
    },
    "logsPolicy": {
        "destination": "CLOUD_LOGGING"
    }
}

Replace the following:

  • PROJECT_ID: the project ID of your project.
  • INSTALL_GPU_DRIVERS: Optional. When set to true, Batch fetches the drivers required for the GPU type that you specify in your Compute Engine VM instance template, and Batch installs them on your behalf. For more information, see how to create a job that uses a GPU.
  • INSTANCE_TEMPLATE_NAME: the name of an existing Compute Engine VM instance template. Learn how to create and list instance templates.

Go

Go

For more information, see the Batch Go API reference documentation.

To authenticate to Batch, set up Application Default Credentials. For more information, see Set up authentication for a local development environment.

import (
	"context"
	"fmt"
	"io"

	batch "cloud.google.com/go/batch/apiv1"
	"cloud.google.com/go/batch/apiv1/batchpb"
	durationpb "google.golang.org/protobuf/types/known/durationpb"
)

// Creates and runs a job that executes the specified script
func createScriptJobWithTemplate(w io.Writer, projectID, region, jobName, templateLink string) error {
	// projectID := "your_project_id"
	// region := "us-central1"
	// jobName := "some-job"
	/* A link to an existing Instance Template. Acceptable formats:
	*  "projects/{project_id}/global/instanceTemplates/{template_name}"
	*  "{template_name}" - if the template is defined in the same project as used to create the Job.
	 */
	// template_link := "my-instance-template"

	ctx := context.Background()
	batchClient, err := batch.NewClient(ctx)
	if err != nil {
		return fmt.Errorf("NewClient: %w", err)
	}
	defer batchClient.Close()

	// Define what will be done as part of the job.
	command := &batchpb.Runnable_Script_Text{
		Text: "echo Hello world! This is task ${BATCH_TASK_INDEX}. This job has a total of ${BATCH_TASK_COUNT} tasks.",
	}

	// We can specify what resources are requested by each task.
	resources := &batchpb.ComputeResource{
		// CpuMilli is milliseconds per cpu-second. This means the task requires 2 whole CPUs.
		CpuMilli:  2000,
		MemoryMib: 16,
	}

	taskSpec := &batchpb.TaskSpec{
		Runnables: []*batchpb.Runnable{{
			Executable: &batchpb.Runnable_Script_{
				Script: &batchpb.Runnable_Script{Command: command},
			},
		}},
		ComputeResource: resources,
		MaxRunDuration: &durationpb.Duration{
			Seconds: 3600,
		},
		MaxRetryCount: 2,
	}

	// Tasks are grouped inside a job using TaskGroups.
	taskGroups := []*batchpb.TaskGroup{
		{
			TaskCount: 4,
			TaskSpec:  taskSpec,
		},
	}

	// Policies are used to define on what kind of virtual machines the tasks will run on.
	// In this case, we are going to use an Instance Template that defines the VM.
	allocationPolicy := &batchpb.AllocationPolicy{
		Instances: []*batchpb.AllocationPolicy_InstancePolicyOrTemplate{{
			PolicyTemplate: &batchpb.AllocationPolicy_InstancePolicyOrTemplate_InstanceTemplate{
				InstanceTemplate: templateLink,
			},
		}},
	}

	// We use Cloud Logging as it's an out of the box available option
	logsPolicy := &batchpb.LogsPolicy{
		Destination: batchpb.LogsPolicy_CLOUD_LOGGING,
	}

	jobLabels := map[string]string{"env": "testing", "type": "script"}

	// The job's parent is the region in which the job will run
	parent := fmt.Sprintf("projects/%s/locations/%s", projectID, region)

	job := batchpb.Job{
		TaskGroups:       taskGroups,
		AllocationPolicy: allocationPolicy,
		Labels:           jobLabels,
		LogsPolicy:       logsPolicy,
	}

	req := &batchpb.CreateJobRequest{
		Parent: parent,
		JobId:  jobName,
		Job:    &job,
	}

	created_job, err := batchClient.CreateJob(ctx, req)
	if err != nil {
		return fmt.Errorf("unable to create job: %w", err)
	}

	fmt.Fprintf(w, "Job created: %v\n", created_job)

	return nil
}

Java

Java

For more information, see the Batch Java API reference documentation.

To authenticate to Batch, set up Application Default Credentials. For more information, see Set up authentication for a local development environment.

import com.google.cloud.batch.v1.AllocationPolicy;
import com.google.cloud.batch.v1.AllocationPolicy.InstancePolicyOrTemplate;
import com.google.cloud.batch.v1.BatchServiceClient;
import com.google.cloud.batch.v1.ComputeResource;
import com.google.cloud.batch.v1.CreateJobRequest;
import com.google.cloud.batch.v1.Job;
import com.google.cloud.batch.v1.LogsPolicy;
import com.google.cloud.batch.v1.LogsPolicy.Destination;
import com.google.cloud.batch.v1.Runnable;
import com.google.cloud.batch.v1.Runnable.Script;
import com.google.cloud.batch.v1.TaskGroup;
import com.google.cloud.batch.v1.TaskSpec;
import com.google.protobuf.Duration;
import java.io.IOException;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;

public class CreateWithTemplate {

  public static void main(String[] args)
      throws IOException, ExecutionException, InterruptedException, TimeoutException {
    // TODO(developer): Replace these variables before running the sample.
    // Project ID or project number of the Cloud project you want to use.
    String projectId = "YOUR_PROJECT_ID";

    // Name of the region you want to use to run the job. Regions that are
    // available for Batch are listed on: https://cloud.google.com/batch/docs/get-started#locations
    String region = "europe-central2";

    // The name of the job that will be created.
    // It needs to be unique for each project and region pair.
    String jobName = "JOB_NAME";

    // A link to an existing Instance Template. Acceptable formats:
    //   * "projects/{projectId}/global/instanceTemplates/{templateName}"
    //   * "{templateName}" - if the template is defined in the same project
    //   as used to create the Job.
    String templateLink = "TEMPLATE_LINK";

    createWithTemplate(projectId, region, jobName, templateLink);
  }

  // This method shows how to create a sample Batch Job that will run
  // a simple command on Cloud Compute instances created using a provided Template.
  public static void createWithTemplate(String projectId, String region, String jobName,
      String templateLink)
      throws IOException, ExecutionException, InterruptedException, TimeoutException {
    // Initialize client that will be used to send requests. This client only needs to be created
    // once, and can be reused for multiple requests. After completing all of your requests, call
    // the `batchServiceClient.close()` method on the client to safely
    // clean up any remaining background resources.
    try (BatchServiceClient batchServiceClient = BatchServiceClient.create()) {

      // Define what will be done as part of the job.
      Runnable runnable =
          Runnable.newBuilder()
              .setScript(
                  Script.newBuilder()
                      .setText(
                          "echo Hello world! This is task ${BATCH_TASK_INDEX}. "
                              + "This job has a total of ${BATCH_TASK_COUNT} tasks.")
                      // You can also run a script from a file. Just remember, that needs to be a
                      // script that's already on the VM that will be running the job.
                      // Using setText() and setPath() is mutually exclusive.
                      // .setPath("/tmp/test.sh")
                      .build())
              .build();

      // We can specify what resources are requested by each task.
      ComputeResource computeResource =
          ComputeResource.newBuilder()
              // In milliseconds per cpu-second. This means the task requires 2 whole CPUs.
              .setCpuMilli(2000)
              // In MiB.
              .setMemoryMib(16)
              .build();

      TaskSpec task =
          TaskSpec.newBuilder()
              // Jobs can be divided into tasks. In this case, we have only one task.
              .addRunnables(runnable)
              .setComputeResource(computeResource)
              .setMaxRetryCount(2)
              .setMaxRunDuration(Duration.newBuilder().setSeconds(3600).build())
              .build();

      // Tasks are grouped inside a job using TaskGroups.
      // Currently, it's possible to have only one task group.
      TaskGroup taskGroup = TaskGroup.newBuilder().setTaskCount(4).setTaskSpec(task).build();

      // Policies are used to define on what kind of virtual machines the tasks will run on.
      // In this case, we tell the system to use an instance template that defines all the
      // required parameters.
      AllocationPolicy allocationPolicy =
          AllocationPolicy.newBuilder()
              .addInstances(
                  InstancePolicyOrTemplate.newBuilder().setInstanceTemplate(templateLink).build())
              .build();

      Job job =
          Job.newBuilder()
              .addTaskGroups(taskGroup)
              .setAllocationPolicy(allocationPolicy)
              .putLabels("env", "testing")
              .putLabels("type", "script")
              // We use Cloud Logging as it's an out of the box available option.
              .setLogsPolicy(
                  LogsPolicy.newBuilder().setDestination(Destination.CLOUD_LOGGING).build())
              .build();

      CreateJobRequest createJobRequest =
          CreateJobRequest.newBuilder()
              // The job's parent is the region in which the job will run.
              .setParent(String.format("projects/%s/locations/%s", projectId, region))
              .setJob(job)
              .setJobId(jobName)
              .build();

      Job result =
          batchServiceClient
              .createJobCallable()
              .futureCall(createJobRequest)
              .get(5, TimeUnit.MINUTES);

      System.out.printf("Successfully created the job: %s", result.getName());
    }
  }
}

Node.js

Node.js

For more information, see the Batch Node.js API reference documentation.

To authenticate to Batch, set up Application Default Credentials. For more information, see Set up authentication for a local development environment.

/**
 * TODO(developer): Uncomment and replace these variables before running the sample.
 */
// const projectId = 'YOUR_PROJECT_ID';
/**
 * The region you want to the job to run in. The regions that support Batch are listed here:
 * https://cloud.google.com/batch/docs/get-started#locations
 */
// const region = 'us-central-1';
/**
 * The name of the job that will be created.
 * It needs to be unique for each project and region pair.
 */
// const jobName = 'YOUR_JOB_NAME';
/**
 * a link to an existing Instance Template. Acceptable formats:
 * "projects/{project_id}/global/instanceTemplates/{template_name}"
 * "{template_name}" - if the template is defined in the same project as used to create the Job.
 */
// const templateLink = 'YOUR_TEMPLATE'

// Imports the Batch library
const batchLib = require('@google-cloud/batch');
const batch = batchLib.protos.google.cloud.batch.v1;

// Instantiates a client
const batchClient = new batchLib.v1.BatchServiceClient();

// Define what will be done as part of the job.
const task = new batch.TaskSpec();
const runnable = new batch.Runnable();
runnable.script = new batch.Runnable.Script();
runnable.script.text =
  'echo Hello world! This is task ${BATCH_TASK_INDEX}. This job has a total of ${BATCH_TASK_COUNT} tasks.';
// You can also run a script from a file. Just remember, that needs to be a script that's
// already on the VM that will be running the job. Using runnable.script.text and runnable.script.path is mutually
// exclusive.
// runnable.script.path = '/tmp/test.sh'
task.runnables = [runnable];

// We can specify what resources are requested by each task.
const resources = new batch.ComputeResource();
resources.cpuMilli = 2000; // in milliseconds per cpu-second. This means the task requires 2 whole CPUs.
resources.memoryMib = 16;
task.computeResource = resources;

task.maxRetryCount = 2;
task.maxRunDuration = {seconds: 3600};

// Tasks are grouped inside a job using TaskGroups.
const group = new batch.TaskGroup();
group.taskCount = 4;
group.taskSpec = task;

// Policies are used to define on what kind of virtual machines the tasks will run on.
// In this case, we tell the system to use "e2-standard-4" machine type.
// Read more about machine types here: https://cloud.google.com/compute/docs/machine-types
const allocationPolicy = new batch.AllocationPolicy();
const instances = new batch.AllocationPolicy.InstancePolicyOrTemplate();
instances.instanceTemplate = templateLink;
allocationPolicy.instances = [instances];

const job = new batch.Job();
job.name = jobName;
job.taskGroups = [group];
job.allocationPolicy = allocationPolicy;
job.labels = {env: 'testing', type: 'script'};
// We use Cloud Logging as it's an option available out of the box
job.logsPolicy = new batch.LogsPolicy();
job.logsPolicy.destination = batch.LogsPolicy.Destination.CLOUD_LOGGING;

// The job's parent is the project and region in which the job will run
const parent = `projects/${projectId}/locations/${region}`;

async function callCreateJob() {
  // Construct request
  const request = {
    parent,
    jobId: jobName,
    job,
  };

  // Run request
  const response = await batchClient.createJob(request);
  console.log(response);
}

await callCreateJob();

Python

Python

For more information, see the Batch Python API reference documentation.

To authenticate to Batch, set up Application Default Credentials. For more information, see Set up authentication for a local development environment.

from google.cloud import batch_v1


def create_script_job_with_template(
    project_id: str, region: str, job_name: str, template_link: str
) -> batch_v1.Job:
    """
    This method shows how to create a sample Batch Job that will run
    a simple command on Cloud Compute instances created using a provided Template.

    Args:
        project_id: project ID or project number of the Cloud project you want to use.
        region: name of the region you want to use to run the job. Regions that are
            available for Batch are listed on: https://cloud.google.com/batch/docs/get-started#locations
        job_name: the name of the job that will be created.
            It needs to be unique for each project and region pair.
        template_link: a link to an existing Instance Template. Acceptable formats:
            * "projects/{project_id}/global/instanceTemplates/{template_name}"
            * "{template_name}" - if the template is defined in the same project as used to create the Job.

    Returns:
        A job object representing the job created.
    """
    client = batch_v1.BatchServiceClient()

    # Define what will be done as part of the job.
    task = batch_v1.TaskSpec()
    runnable = batch_v1.Runnable()
    runnable.script = batch_v1.Runnable.Script()
    runnable.script.text = "echo Hello world! This is task ${BATCH_TASK_INDEX}. This job has a total of ${BATCH_TASK_COUNT} tasks."
    # You can also run a script from a file. Just remember, that needs to be a script that's
    # already on the VM that will be running the job. Using runnable.script.text and runnable.script.path is mutually
    # exclusive.
    # runnable.script.path = '/tmp/test.sh'
    task.runnables = [runnable]

    # We can specify what resources are requested by each task.
    resources = batch_v1.ComputeResource()
    resources.cpu_milli = 2000  # in milliseconds per cpu-second. This means the task requires 2 whole CPUs.
    resources.memory_mib = 16
    task.compute_resource = resources

    task.max_retry_count = 2
    task.max_run_duration = "3600s"

    # Tasks are grouped inside a job using TaskGroups.
    # Currently, it's possible to have only one task group.
    group = batch_v1.TaskGroup()
    group.task_count = 4
    group.task_spec = task

    # Policies are used to define on what kind of virtual machines the tasks will run on.
    # In this case, we tell the system to use an instance template that defines all the
    # required parameters.
    allocation_policy = batch_v1.AllocationPolicy()
    instances = batch_v1.AllocationPolicy.InstancePolicyOrTemplate()
    instances.instance_template = template_link
    allocation_policy.instances = [instances]

    job = batch_v1.Job()
    job.task_groups = [group]
    job.allocation_policy = allocation_policy
    job.labels = {"env": "testing", "type": "script"}
    # We use Cloud Logging as it's an out of the box available option
    job.logs_policy = batch_v1.LogsPolicy()
    job.logs_policy.destination = batch_v1.LogsPolicy.Destination.CLOUD_LOGGING

    create_request = batch_v1.CreateJobRequest()
    create_request.job = job
    create_request.job_id = job_name
    # The job's parent is the region in which the job will run
    create_request.parent = f"projects/{project_id}/locations/{region}"

    return client.create_job(create_request)

C++

C++

For more information, see the Batch C++ API reference documentation.

To authenticate to Batch, set up Application Default Credentials. For more information, see Set up authentication for a local development environment.

#include "google/cloud/batch/v1/batch_client.h"

  [](std::string const& project_id, std::string const& location_id,
     std::string const& job_id, std::string const& template_name) {
    // Initialize the request; start with the fields that depend on the sample
    // input.
    google::cloud::batch::v1::CreateJobRequest request;
    request.set_parent("projects/" + project_id + "/locations/" + location_id);
    request.set_job_id(job_id);
    // Most of the job description is fixed in this example; use a string to
    // initialize it, and then override the template name.
    auto constexpr kText = R"pb(
      task_groups {
        task_count: 4
        task_spec {
          compute_resource { cpu_milli: 500 memory_mib: 16 }
          max_retry_count: 2
          max_run_duration { seconds: 3600 }
          runnables {
            script {
              text: "echo Hello world! This is task ${BATCH_TASK_INDEX}. This job has a total of ${BATCH_TASK_COUNT} tasks."
            }
          }
        }
      }
      labels { key: "env" value: "testing" }
      labels { key: "type" value: "script" }
      logs_policy { destination: CLOUD_LOGGING }
    )pb";
    auto* job = request.mutable_job();
    if (!google::protobuf::TextFormat::ParseFromString(kText, job)) {
      throw std::runtime_error("Error parsing Job description");
    }
    job->mutable_allocation_policy()->add_instances()->set_instance_template(
        template_name);
    // Create a client and issue the request.
    auto client = google::cloud::batch_v1::BatchServiceClient(
        google::cloud::batch_v1::MakeBatchServiceConnection());
    auto response = client.CreateJob(request);
    if (!response) throw std::move(response).status();
    std::cout << "Job : " << response->DebugString() << "\n";
  }

What's next