Examples with SparkConf org.apache.spark.SparkConf used on opensource projects

Example 61 with SparkConf

use of org.apache.spark.SparkConf in project deeplearning4j by deeplearning4j.

the class ExportSupportTest method testLocalSupported.

@Test
public void testLocalSupported() throws IOException {
    assertSupported(new SparkConf().setMaster("local").set(FS_CONF, "file:///"));
    assertSupported(new SparkConf().setMaster("local[2]").set(FS_CONF, "file:///"));
    assertSupported(new SparkConf().setMaster("local[64]").set(FS_CONF, "file:///"));
    assertSupported(new SparkConf().setMaster("local[*]").set(FS_CONF, "file:///"));
    assertSupported(new SparkConf().setMaster("local").set(FS_CONF, "hdfs://localhost:9000"));
}

Example 62 with SparkConf

use of org.apache.spark.SparkConf in project Gaffer by gchq.

the class ImportJavaRDDOfElementsHandlerTest method checkImportJavaRDDOfElements.

@Test
public void checkImportJavaRDDOfElements() throws OperationException, IOException, InterruptedException {
    final Graph graph1 = new Graph.Builder().addSchema(getClass().getResourceAsStream("/schema/dataSchema.json")).addSchema(getClass().getResourceAsStream("/schema/dataTypes.json")).addSchema(getClass().getResourceAsStream("/schema/storeTypes.json")).storeProperties(getClass().getResourceAsStream("/store.properties")).build();
    final List<Element> elements = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        final Entity entity = new Entity(TestGroups.ENTITY);
        entity.setVertex("" + i);
        final Edge edge1 = new Edge(TestGroups.EDGE);
        edge1.setSource("" + i);
        edge1.setDestination("B");
        edge1.setDirected(false);
        edge1.putProperty(TestPropertyNames.COUNT, 2);
        final Edge edge2 = new Edge(TestGroups.EDGE);
        edge2.setSource("" + i);
        edge2.setDestination("C");
        edge2.setDirected(false);
        edge2.putProperty(TestPropertyNames.COUNT, 4);
        elements.add(edge1);
        elements.add(edge2);
        elements.add(entity);
    }
    final User user = new User();
    final SparkConf sparkConf = new SparkConf().setMaster("local").setAppName("testCheckGetCorrectElementsInJavaRDDForEntitySeed").set("spark.serializer", "org.apache.spark.serializer.KryoSerializer").set("spark.kryo.registrator", "uk.gov.gchq.gaffer.spark.serialisation.kryo.Registrator").set("spark.driver.allowMultipleContexts", "true");
    final JavaSparkContext sparkContext = new JavaSparkContext(sparkConf);
    // Create Hadoop configuration and serialise to a string
    final Configuration configuration = new Configuration();
    final ByteArrayOutputStream baos = new ByteArrayOutputStream();
    configuration.write(new DataOutputStream(baos));
    final String configurationString = new String(baos.toByteArray(), CommonConstants.UTF_8);
    final String outputPath = this.getClass().getResource("/").getPath().toString() + "load";
    final String failurePath = this.getClass().getResource("/").getPath().toString() + "failure";
    final File file = new File(outputPath);
    if (file.exists()) {
        FileUtils.forceDelete(file);
    }
    final JavaRDD<Element> elementJavaRDD = sparkContext.parallelize(elements);
    final ImportJavaRDDOfElements addRdd = new ImportJavaRDDOfElements.Builder().javaSparkContext(sparkContext).input(elementJavaRDD).option("outputPath", outputPath).option("failurePath", failurePath).build();
    graph1.execute(addRdd, user);
    FileUtils.forceDelete(file);
    // Check all elements were added
    final GetJavaRDDOfAllElements rddQuery = new GetJavaRDDOfAllElements.Builder().javaSparkContext(sparkContext).option(AbstractGetRDDHandler.HADOOP_CONFIGURATION_KEY, configurationString).build();
    final JavaRDD<Element> rdd = graph1.execute(rddQuery, user);
    if (rdd == null) {
        fail("No RDD returned");
    }
    final Set<Element> results = new HashSet<>(rdd.collect());
    assertEquals(elements.size(), results.size());
    sparkContext.stop();
}

Example 63 with SparkConf

use of org.apache.spark.SparkConf in project Gaffer by gchq.

the class ImportKeyValueJavaPairRDDToAccumuloHandlerTest method checkImportKeyValueJavaPairRDD.

@Test
public void checkImportKeyValueJavaPairRDD() throws OperationException, IOException, InterruptedException {
    final Graph graph1 = new Graph.Builder().addSchema(getClass().getResourceAsStream("/schema/dataSchema.json")).addSchema(getClass().getResourceAsStream("/schema/dataTypes.json")).addSchema(getClass().getResourceAsStream("/schema/storeSchema.json")).addSchema(getClass().getResourceAsStream("/schema/storeTypes.json")).storeProperties(getClass().getResourceAsStream("/store.properties")).build();
    final List<Element> elements = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        final Entity entity = new Entity(TestGroups.ENTITY);
        entity.setVertex("" + i);
        final Edge edge1 = new Edge(TestGroups.EDGE);
        edge1.setSource("" + i);
        edge1.setDestination("B");
        edge1.setDirected(false);
        edge1.putProperty(TestPropertyNames.COUNT, 2);
        final Edge edge2 = new Edge(TestGroups.EDGE);
        edge2.setSource("" + i);
        edge2.setDestination("C");
        edge2.setDirected(false);
        edge2.putProperty(TestPropertyNames.COUNT, 4);
        elements.add(edge1);
        elements.add(edge2);
        elements.add(entity);
    }
    final User user = new User();
    final SparkConf sparkConf = new SparkConf().setMaster("local").setAppName("testCheckGetCorrectElementsInJavaRDDForEntitySeed").set("spark.serializer", "org.apache.spark.serializer.KryoSerializer").set("spark.kryo.registrator", "uk.gov.gchq.gaffer.spark.serialisation.kryo.Registrator").set("spark.driver.allowMultipleContexts", "true");
    final JavaSparkContext sparkContext = new JavaSparkContext(sparkConf);
    // Create Hadoop configuration and serialise to a string
    final Configuration configuration = new Configuration();
    final ByteArrayOutputStream baos = new ByteArrayOutputStream();
    configuration.write(new DataOutputStream(baos));
    final String configurationString = new String(baos.toByteArray(), CommonConstants.UTF_8);
    final String outputPath = this.getClass().getResource("/").getPath().toString() + "load";
    final String failurePath = this.getClass().getResource("/").getPath().toString() + "failure";
    final File file = new File(outputPath);
    if (file.exists()) {
        FileUtils.forceDelete(file);
    }
    final ElementConverterFunction func = new ElementConverterFunction(sparkContext.broadcast(new ByteEntityAccumuloElementConverter(graph1.getSchema())));
    final JavaPairRDD<Key, Value> elementJavaRDD = sparkContext.parallelize(elements).flatMapToPair(func);
    final ImportKeyValueJavaPairRDDToAccumulo addRdd = new ImportKeyValueJavaPairRDDToAccumulo.Builder().input(elementJavaRDD).outputPath(outputPath).failurePath(failurePath).build();
    graph1.execute(addRdd, user);
    FileUtils.forceDelete(file);
    // Check all elements were added
    final GetJavaRDDOfAllElements rddQuery = new GetJavaRDDOfAllElements.Builder().javaSparkContext(sparkContext).option(AbstractGetRDDHandler.HADOOP_CONFIGURATION_KEY, configurationString).build();
    final JavaRDD<Element> rdd = graph1.execute(rddQuery, user);
    if (rdd == null) {
        fail("No RDD returned");
    }
    final Set<Element> results = new HashSet<>(rdd.collect());
    assertEquals(elements.size(), results.size());
    sparkContext.stop();
}

Example 64 with SparkConf

use of org.apache.spark.SparkConf in project Gaffer by gchq.

the class GetDataFrameOfElementsExample method runExamples.

@Override
public void runExamples() {
    // Need to actively turn logging on and off as needed as Spark produces some logs
    // even when the log level is set to off.
    ROOT_LOGGER.setLevel(Level.OFF);
    final SparkConf sparkConf = new SparkConf().setMaster("local").setAppName("getDataFrameOfElementsWithEntityGroup").set("spark.serializer", "org.apache.spark.serializer.KryoSerializer").set("spark.kryo.registrator", "uk.gov.gchq.gaffer.spark.serialisation.kryo.Registrator").set("spark.driver.allowMultipleContexts", "true");
    final SparkContext sc = new SparkContext(sparkConf);
    sc.setLogLevel("OFF");
    final SQLContext sqlc = new SQLContext(sc);
    final Graph graph = getGraph();
    try {
        getDataFrameOfElementsWithEntityGroup(sqlc, graph);
        getDataFrameOfElementsWithEdgeGroup(sqlc, graph);
    } catch (final OperationException e) {
        throw new RuntimeException(e);
    }
    sc.stop();
    ROOT_LOGGER.setLevel(Level.INFO);
}

Example 65 with SparkConf

use of org.apache.spark.SparkConf in project Gaffer by gchq.

the class GetJavaRDDOfAllElementsExample method runExamples.

@Override
public void runExamples() {
    // Need to actively turn logging on and off as needed as Spark produces some logs
    // even when the log level is set to off.
    ROOT_LOGGER.setLevel(Level.OFF);
    final SparkConf sparkConf = new SparkConf().setMaster("local").setAppName("GetJavaRDDOfAllElementsExample").set("spark.serializer", "org.apache.spark.serializer.KryoSerializer").set("spark.kryo.registrator", "uk.gov.gchq.gaffer.spark.serialisation.kryo.Registrator").set("spark.driver.allowMultipleContexts", "true");
    final JavaSparkContext sc = new JavaSparkContext(sparkConf);
    sc.setLogLevel("OFF");
    final Graph graph = getGraph();
    try {
        getJavaRddOfAllElements(sc, graph);
        getJavaRddOfAllElementsReturningEdgesOnly(sc, graph);
    } catch (final OperationException e) {
        throw new RuntimeException(e);
    }
    sc.stop();
    ROOT_LOGGER.setLevel(Level.INFO);
}