com.amazonaws.services.elasticmapreduce.model.Application Java Examples

private Cluster makeCluster(String id)
{
    return new Cluster()
            .withId(id)
            .withName("name")
            .withAutoScalingRole("autoscaling_role")
            .withCustomAmiId("custom_ami")
            .withInstanceCollectionType("instance_collection_type")
            .withLogUri("log_uri")
            .withMasterPublicDnsName("master_public_dns")
            .withReleaseLabel("release_label")
            .withRunningAmiVersion("running_ami")
            .withScaleDownBehavior("scale_down_behavior")
            .withServiceRole("service_role")
            .withApplications(new Application().withName("name").withVersion("version"))
            .withTags(new Tag("key", "value"));
}
 

/**
 * Converts the given list of {@link EmrClusterDefinitionApplication} into a list of {@link Application}
 *
 * @param emrClusterDefinitionApplications list of {@link EmrClusterDefinitionApplication}
 *
 * @return list {@link Application}
 */
public List<Application> getApplications(List<EmrClusterDefinitionApplication> emrClusterDefinitionApplications)
{
    List<Application> applications = new ArrayList<>();
    for (EmrClusterDefinitionApplication emrClusterDefinitionApplication : emrClusterDefinitionApplications)
    {
        Application application = new Application();
        application.setName(emrClusterDefinitionApplication.getName());
        application.setVersion(emrClusterDefinitionApplication.getVersion());
        application.setArgs(emrClusterDefinitionApplication.getArgs());

        List<Parameter> additionalInfoList = emrClusterDefinitionApplication.getAdditionalInfoList();
        if (!CollectionUtils.isEmpty(additionalInfoList))
        {
            application.setAdditionalInfo(getMap(additionalInfoList));
        }

        applications.add(application);
    }
    return applications;
}
 

@Test
public void test()
        throws Exception
{
    RunJobFlowRequest request = new RunJobFlowRequest()
            .withName("Digdag Test")
            .withReleaseLabel("emr-5.2.0")
            .withApplications(Stream.of("Hadoop", "Hive", "Spark", "Flink")
                    .map(s -> new Application().withName(s))
                    .collect(toList()))
            .withJobFlowRole("EMR_EC2_DefaultRole")
            .withServiceRole("EMR_DefaultRole")
            .withVisibleToAllUsers(true)
            .withLogUri(tmpS3FolderUri + "/logs/")
            .withInstances(new JobFlowInstancesConfig()
                    .withEc2KeyName("digdag-test")
                    .withInstanceCount(1)
                    .withKeepJobFlowAliveWhenNoSteps(true)
                    .withMasterInstanceType("m3.xlarge")
                    .withSlaveInstanceType("m3.xlarge"));

    RunJobFlowResult result = emr.runJobFlow(request);

    String clusterId = result.getJobFlowId();

    clusterIds.add(clusterId);

    Id attemptId = pushAndStart(server.endpoint(), projectDir, "emr", ImmutableMap.of(
            "test_s3_folder", tmpS3FolderUri.toString(),
            "test_cluster", clusterId,
            "outfile", outfile.toString()));
    expect(Duration.ofMinutes(30), attemptSuccess(server.endpoint(), attemptId));

    validateTdSparkQueryOutput();

    assertThat(Files.exists(outfile), is(true));
}
 

protected String fireEMRJob(String paramsStr,String clusterId){
	StepFactory stepFactory = new StepFactory();
	AmazonElasticMapReduceClient emr = new AmazonElasticMapReduceClient();
	emr.setRegion(Region.getRegion(Regions.fromName(System.getenv().get("AWS_REGION"))));
	Application sparkConfig = new Application()
			.withName("Spark");
	
	String[] params = paramsStr.split(",");
	StepConfig enabledebugging = new StepConfig()
			.withName("Enable debugging")
			.withActionOnFailure("TERMINATE_JOB_FLOW")
			.withHadoopJarStep(stepFactory.newEnableDebuggingStep());
	
	HadoopJarStepConfig sparkStepConf = new HadoopJarStepConfig()
			.withJar("command-runner.jar")
			.withArgs(params);	
	
	final StepConfig sparkStep = new StepConfig()
			.withName("Spark Step")
			.withActionOnFailure("CONTINUE")
			.withHadoopJarStep(sparkStepConf);

	
	AddJobFlowStepsRequest request = new AddJobFlowStepsRequest(clusterId)
			.withSteps(new ArrayList<StepConfig>(){{add(sparkStep);}});
			

	AddJobFlowStepsResult result = emr.addJobFlowSteps(request);
	return result.getStepIds().get(0);
}
 

private NewCluster submitNewClusterRequest(AmazonElasticMapReduce emr, String tag, StepCompiler stepCompiler,
        Config cluster, Filer filer, ParameterCompiler parameterCompiler)
        throws IOException
{
    RemoteFile runner = prepareRunner(filer, tag);

    // Compile steps
    stepCompiler.compile(runner);

    List<StepConfig> stepConfigs = stepCompiler.stepConfigs();

    Config ec2 = cluster.getNested("ec2");
    Config master = ec2.getNestedOrGetEmpty("master");
    List<Config> core = ec2.getOptional("core", Config.class).transform(ImmutableList::of).or(ImmutableList.of());
    List<Config> task = ec2.getListOrEmpty("task", Config.class);

    List<String> applications = cluster.getListOrEmpty("applications", String.class);
    if (applications.isEmpty()) {
        applications = ImmutableList.of("Hadoop", "Hive", "Spark", "Flink");
    }

    // TODO: allow configuring additional application parameters
    List<Application> applicationConfigs = applications.stream()
            .map(application -> new Application().withName(application))
            .collect(toList());

    // TODO: merge configurations with the same classification?
    List<Configuration> configurations = cluster.getListOrEmpty("configurations", JsonNode.class).stream()
            .map(this::configurations)
            .flatMap(Collection::stream)
            .collect(toList());

    List<JsonNode> bootstrap = cluster.getListOrEmpty("bootstrap", JsonNode.class);
    List<BootstrapActionConfig> bootstrapActions = new ArrayList<>();
    for (int i = 0; i < bootstrap.size(); i++) {
        bootstrapActions.add(bootstrapAction(i + 1, bootstrap.get(i), tag, filer, runner, parameterCompiler));
    }

    // Stage files to S3
    filer.stageFiles();

    Optional<String> subnetId = ec2.getOptional("subnet_id", String.class);

    String defaultMasterInstanceType;
    String defaultCoreInstanceType;
    String defaultTaskInstanceType;

    if (subnetId.isPresent()) {
        // m4 requires VPC (subnet id)
        defaultMasterInstanceType = "m4.2xlarge";
        defaultCoreInstanceType = "m4.xlarge";
        defaultTaskInstanceType = "m4.xlarge";
    }
    else {
        defaultMasterInstanceType = "m3.2xlarge";
        defaultCoreInstanceType = "m3.xlarge";
        defaultTaskInstanceType = "m3.xlarge";
    }

    RunJobFlowRequest request = new RunJobFlowRequest()
            .withName(cluster.get("name", String.class, "Digdag") + " (" + tag + ")")
            .withReleaseLabel(cluster.get("release", String.class, "emr-5.2.0"))
            .withSteps(stepConfigs)
            .withBootstrapActions(bootstrapActions)
            .withApplications(applicationConfigs)
            .withLogUri(cluster.get("logs", String.class, null))
            .withJobFlowRole(cluster.get("cluster_role", String.class, "EMR_EC2_DefaultRole"))
            .withServiceRole(cluster.get("service_role", String.class, "EMR_DefaultRole"))
            .withTags(new Tag().withKey("DIGDAG_CLUSTER_ID").withValue(tag))
            .withVisibleToAllUsers(cluster.get("visible", boolean.class, true))
            .withConfigurations(configurations)
            .withInstances(new JobFlowInstancesConfig()
                    .withInstanceGroups(ImmutableList.<InstanceGroupConfig>builder()
                            // Master Node
                            .add(instanceGroupConfig("Master", master, "MASTER", defaultMasterInstanceType, 1))
                            // Core Group
                            .addAll(instanceGroupConfigs("Core", core, "CORE", defaultCoreInstanceType))
                            // Task Groups
                            .addAll(instanceGroupConfigs("Task %d", task, "TASK", defaultTaskInstanceType))
                            .build()
                    )
                    .withAdditionalMasterSecurityGroups(ec2.getListOrEmpty("additional_master_security_groups", String.class))
                    .withAdditionalSlaveSecurityGroups(ec2.getListOrEmpty("additional_slave_security_groups", String.class))
                    .withEmrManagedMasterSecurityGroup(ec2.get("emr_managed_master_security_group", String.class, null))
                    .withEmrManagedSlaveSecurityGroup(ec2.get("emr_managed_slave_security_group", String.class, null))
                    .withServiceAccessSecurityGroup(ec2.get("service_access_security_group", String.class, null))
                    .withTerminationProtected(cluster.get("termination_protected", boolean.class, false))
                    .withPlacement(cluster.getOptional("availability_zone", String.class)
                            .transform(zone -> new PlacementType().withAvailabilityZone(zone)).orNull())
                    .withEc2SubnetId(subnetId.orNull())
                    .withEc2KeyName(ec2.get("key", String.class))
                    .withKeepJobFlowAliveWhenNoSteps(!cluster.get("auto_terminate", boolean.class, true)));

    logger.info("Submitting EMR job with {} steps(s)", request.getSteps().size());
    RunJobFlowResult result = emr.runJobFlow(request);
    logger.info("Submitted EMR job with {} step(s): {}", request.getSteps().size(), result.getJobFlowId(), result);

    return NewCluster.of(result.getJobFlowId(), request.getSteps().size());
}