AWS-SDK-Cross-Api

This SDK try to simplify and get a standard way to traits cluster between emr and data pipeline. Is easy to create a pipeline programmatically using the Objects provided.

Models to compose

• AWSAction
• AWSActivity
• AWSConfiguration
• AWSDatabase
• AWSDataFormat
• AWSDatanode
• AWSPrecondiction
• AWSProperty
• AWSResource
• AWSSChedule

EMR Cluster Sample

how to create en EMR Cluster for data pipeline service in the traditional way

    String name = "EmrClusterForBackup";
    String id = "EmrClusterForBackup";

    Field type = new Field().withKey("type").withStringValue("EmrCluster");
    Field amiVersion = new Field().withKey("amiVersion").withStringValue("3.10.0");
    Field masterInstanceType = new Field().withKey("masterInstanceType").withStringValue("m3.xlarge");
    Field coreInstanceType = new Field().withKey("coreInstanceType").withStringValue("m3.xlarge");
    Field coreInstanceCount = new Field().withKey("coreInstanceCount").withStringValue("1");
    Field region = new Field().withKey("region").withStringValue("#{myDDBRegion}");
    Field terminateAfter = new Field().withKey("terminateAfter").withStringValue("12 hours");
    Field bootstrapAction = new Field().withKey("bootstrapAction").withStringValue("s3://elasticmapreduce" +
            "/bootstrap-actions/configure-hadoop, --yarn-key-value,yarn.nodemanager.resource.memory-mb=11520," +
            "--yarn-key-value,yarn.scheduler.maximum-allocation-mb=11520," +
            "--yarn-key-value,yarn.scheduler.minimum-allocation-mb=1440," +
            "--yarn-key-value,yarn.app.mapreduce.am.resource.mb=2880," +
            "--mapred-key-value,mapreduce.map.memory.mb=5760," +
            "--mapred-key-value,mapreduce.map.java.opts=-Xmx4608M," +
            "--mapred-key-value,mapreduce.reduce.memory.mb=2880," +
            "--mapred-key-value,mapreduce.reduce.java.opts=-Xmx2304m," +
            "--mapred-key-value,mapreduce.map.speculative=false");

    List<Field> fieldsList = Lists.newArrayList(type, amiVersion, masterInstanceType,
            coreInstanceCount, coreInstanceType, region, terminateAfter, bootstrapAction);

    return new PipelineObject().withName(name).withId(id).withFields(fieldsList);

using this SDK

    MyCluster.MyClusterBuilder.newOne()
            .withClusterHwConf( ClusterHwConfBuilder
                                               .newOne()
                                               .withCoreInstanceType(new MyString("m3.xlarge"))
                                               .withMasterInstanceType(new MyString("m3.xlarge"))
                                               .withCoreInstanceCount(new MyInteger(2))
                                               .build())
            .withClusterSecurityConf(secConf)
            .withClusterBehaviourConf( ClusterBehaviourConfBuilder.newOne()
                                                      .withTerminateAfter(new MyString("12 Hours"))
                                                      .withBootstrapAction("ecc")
                                                      .build();)
            .withClusterSwConf(swConf)
            .build("EmrClusterForBackup", "EmrClusterForBackup");

Pipeline definition example

    MyPipelineService myPipelineService = new MyPipelineService( DataPipelineClientBuilder.defaultClient());


    //DEFINE CLUSTER
    EmrCluster myCluster = getEMRCluster(); //as we described above


    //DEFINE STEP
    JavaSparkStep step = new JavaSparkStep.JavaSparkBuilder().
            withMasterUrl("yarn-cluster", SparkDeployMode.cluster)
            .withJar("myApp.jar")
            .withMainClass("com.ciao.MainClass")
            .create();


    //DEFINE ACTIVITY
    EmrActivity myActivity = MyActivity.newEMR()
            .withStep(new MyString(step.toString()))
            .withTag("MyActivityExample", "MyActivityExample")
            .withRunOn(myCluster).creation();


    //DEFINE PIPELINE
    MyPipeline pipeline = MyPipeline.newOne(provideEmptyDefault())
            .addActivity(myActivity)
            .create();



    String pipelineId = myPipelineService.createAndDefine("prova", pipeline, new HashMap<>())

Todo

• emr cluster builder adding control on not null fields (role is required)

• myPipelineService.createAndDefine(s3Key, myPipeline, null); required not null on parameters

• when you define Default object the id must be Default