Examples with ETLBatchConfig co.cask.cdap.etl.proto.v2.ETLBatchConfig used on opensource projects

Example 41 with ETLBatchConfig

use of co.cask.cdap.etl.proto.v2.ETLBatchConfig in project cdap by caskdata.

the class DataPipelineTest method testMultipleJoiner.

private void testMultipleJoiner(Engine engine) throws Exception {
    /*
     * source1 ----> t1 ------
     *                        | --> innerjoin ----> t4 ------
     * source2 ----> t2 ------                                 |
     *                                                         | ---> outerjoin --> sink1
     *                                                         |
     * source3 -------------------- t3 ------------------------
     */
    Schema inputSchema1 = Schema.recordOf("customerRecord", Schema.Field.of("customer_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("customer_name", Schema.of(Schema.Type.STRING)));
    Schema inputSchema2 = Schema.recordOf("itemRecord", Schema.Field.of("item_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("item_price", Schema.of(Schema.Type.LONG)), Schema.Field.of("cust_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("cust_name", Schema.of(Schema.Type.STRING)));
    Schema inputSchema3 = Schema.recordOf("transactionRecord", Schema.Field.of("t_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("c_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("i_id", Schema.of(Schema.Type.STRING)));
    Schema outSchema2 = Schema.recordOf("join.output", Schema.Field.of("t_id", Schema.nullableOf(Schema.of(Schema.Type.STRING))), Schema.Field.of("c_id", Schema.nullableOf(Schema.of(Schema.Type.STRING))), Schema.Field.of("i_id", Schema.nullableOf(Schema.of(Schema.Type.STRING))), Schema.Field.of("customer_id", Schema.nullableOf(Schema.of(Schema.Type.STRING))), Schema.Field.of("customer_name", Schema.nullableOf(Schema.of(Schema.Type.STRING))), Schema.Field.of("item_id", Schema.nullableOf(Schema.of(Schema.Type.STRING))), Schema.Field.of("item_price", Schema.nullableOf(Schema.of(Schema.Type.LONG))), Schema.Field.of("cust_id", Schema.nullableOf(Schema.of(Schema.Type.STRING))), Schema.Field.of("cust_name", Schema.nullableOf(Schema.of(Schema.Type.STRING))));
    String source1MulitJoinInput = "multiJoinSource1-" + engine;
    String source2MultiJoinInput = "multiJoinSource2-" + engine;
    String source3MultiJoinInput = "multiJoinSource3-" + engine;
    String outputName = "multiJoinOutput-" + engine;
    String sinkName = "multiJoinOutputSink-" + engine;
    String outerJoinName = "multiJoinOuter-" + engine;
    ETLBatchConfig etlConfig = ETLBatchConfig.builder("* * * * *").addStage(new ETLStage("source1", MockSource.getPlugin(source1MulitJoinInput, inputSchema1))).addStage(new ETLStage("source2", MockSource.getPlugin(source2MultiJoinInput, inputSchema2))).addStage(new ETLStage("source3", MockSource.getPlugin(source3MultiJoinInput, inputSchema3))).addStage(new ETLStage("t1", IdentityTransform.getPlugin())).addStage(new ETLStage("t2", IdentityTransform.getPlugin())).addStage(new ETLStage("t3", IdentityTransform.getPlugin())).addStage(new ETLStage("t4", IdentityTransform.getPlugin())).addStage(new ETLStage("innerjoin", MockJoiner.getPlugin("t1.customer_id=t2.cust_id", "t1,t2", ""))).addStage(new ETLStage(outerJoinName, MockJoiner.getPlugin("t4.item_id=t3.i_id", "", ""))).addStage(new ETLStage(sinkName, MockSink.getPlugin(outputName))).addConnection("source1", "t1").addConnection("source2", "t2").addConnection("source3", "t3").addConnection("t1", "innerjoin").addConnection("t2", "innerjoin").addConnection("innerjoin", "t4").addConnection("t3", outerJoinName).addConnection("t4", outerJoinName).addConnection(outerJoinName, sinkName).setEngine(engine).build();
    AppRequest<ETLBatchConfig> appRequest = new AppRequest<>(APP_ARTIFACT, etlConfig);
    ApplicationId appId = NamespaceId.DEFAULT.app("JoinerApp-" + engine);
    ApplicationManager appManager = deployApplication(appId.toId(), appRequest);
    StructuredRecord recordSamuel = StructuredRecord.builder(inputSchema1).set("customer_id", "1").set("customer_name", "samuel").build();
    StructuredRecord recordBob = StructuredRecord.builder(inputSchema1).set("customer_id", "2").set("customer_name", "bob").build();
    StructuredRecord recordJane = StructuredRecord.builder(inputSchema1).set("customer_id", "3").set("customer_name", "jane").build();
    StructuredRecord recordCar = StructuredRecord.builder(inputSchema2).set("item_id", "11").set("item_price", 10000L).set("cust_id", "1").set("cust_name", "samuel").build();
    StructuredRecord recordBike = StructuredRecord.builder(inputSchema2).set("item_id", "22").set("item_price", 100L).set("cust_id", "3").set("cust_name", "jane").build();
    StructuredRecord recordTrasCar = StructuredRecord.builder(inputSchema3).set("t_id", "1").set("c_id", "1").set("i_id", "11").build();
    StructuredRecord recordTrasBike = StructuredRecord.builder(inputSchema3).set("t_id", "2").set("c_id", "3").set("i_id", "22").build();
    StructuredRecord recordTrasPlane = StructuredRecord.builder(inputSchema3).set("t_id", "3").set("c_id", "4").set("i_id", "33").build();
    // write one record to each source
    DataSetManager<Table> inputManager = getDataset(NamespaceId.DEFAULT.dataset(source1MulitJoinInput));
    MockSource.writeInput(inputManager, ImmutableList.of(recordSamuel, recordBob, recordJane));
    inputManager = getDataset(NamespaceId.DEFAULT.dataset(source2MultiJoinInput));
    MockSource.writeInput(inputManager, ImmutableList.of(recordCar, recordBike));
    inputManager = getDataset(NamespaceId.DEFAULT.dataset(source3MultiJoinInput));
    MockSource.writeInput(inputManager, ImmutableList.of(recordTrasCar, recordTrasBike, recordTrasPlane));
    WorkflowManager workflowManager = appManager.getWorkflowManager(SmartWorkflow.NAME);
    workflowManager.start();
    workflowManager.waitForRun(ProgramRunStatus.COMPLETED, 5, TimeUnit.MINUTES);
    StructuredRecord joinRecordSamuel = StructuredRecord.builder(outSchema2).set("customer_id", "1").set("customer_name", "samuel").set("item_id", "11").set("item_price", 10000L).set("cust_id", "1").set("cust_name", "samuel").set("t_id", "1").set("c_id", "1").set("i_id", "11").build();
    StructuredRecord joinRecordJane = StructuredRecord.builder(outSchema2).set("customer_id", "3").set("customer_name", "jane").set("item_id", "22").set("item_price", 100L).set("cust_id", "3").set("cust_name", "jane").set("t_id", "2").set("c_id", "3").set("i_id", "22").build();
    StructuredRecord joinRecordPlane = StructuredRecord.builder(outSchema2).set("t_id", "3").set("c_id", "4").set("i_id", "33").build();
    DataSetManager<Table> sinkManager = getDataset(outputName);
    Set<StructuredRecord> expected = ImmutableSet.of(joinRecordSamuel, joinRecordJane, joinRecordPlane);
    Set<StructuredRecord> actual = Sets.newHashSet(MockSink.readOutput(sinkManager));
    Assert.assertEquals(expected, actual);
    validateMetric(3, appId, outerJoinName + ".records.out");
    validateMetric(3, appId, sinkName + ".records.in");
}

Example 42 with ETLBatchConfig

use of co.cask.cdap.etl.proto.v2.ETLBatchConfig in project cdap by caskdata.

the class DataPipelineTest method testMacroEvaluationActionPipeline.

public void testMacroEvaluationActionPipeline(Engine engine) throws Exception {
    ETLStage action1 = new ETLStage("action1", MockAction.getPlugin("actionTable", "action1.row", "action1.column", "${value}"));
    ETLBatchConfig etlConfig = co.cask.cdap.etl.proto.v2.ETLBatchConfig.builder("* * * * *").addStage(action1).setEngine(engine).build();
    // set runtime arguments for macro substitution
    Map<String, String> runtimeArguments = ImmutableMap.of("value", "macroValue");
    AppRequest<co.cask.cdap.etl.proto.v2.ETLBatchConfig> appRequest = new AppRequest<>(APP_ARTIFACT, etlConfig);
    ApplicationId appId = NamespaceId.DEFAULT.app("macroActionTest-" + engine);
    ApplicationManager appManager = deployApplication(appId, appRequest);
    WorkflowManager manager = appManager.getWorkflowManager(SmartWorkflow.NAME);
    manager.setRuntimeArgs(runtimeArguments);
    manager.start(ImmutableMap.of("logical.start.time", "0"));
    manager.waitForRun(ProgramRunStatus.COMPLETED, 3, TimeUnit.MINUTES);
    DataSetManager<Table> actionTableDS = getDataset("actionTable");
    Assert.assertEquals("macroValue", MockAction.readOutput(actionTableDS, "action1.row", "action1.column"));
    appManager.getHistory(appId.workflow(SmartWorkflow.NAME).toId(), ProgramRunStatus.FAILED);
}

Example 43 with ETLBatchConfig

use of co.cask.cdap.etl.proto.v2.ETLBatchConfig in project cdap by caskdata.

the class DataPipelineApp method configure.

@Override
public void configure() {
    ETLBatchConfig config = getConfig();
    setDescription(DEFAULT_DESCRIPTION);
    PipelineSpecGenerator<ETLBatchConfig, BatchPipelineSpec> specGenerator = new BatchPipelineSpecGenerator(getConfigurer(), ImmutableSet.of(BatchSource.PLUGIN_TYPE), ImmutableSet.of(BatchSink.PLUGIN_TYPE, SparkSink.PLUGIN_TYPE), TimePartitionedFileSet.class, FileSetProperties.builder().setInputFormat(AvroKeyInputFormat.class).setOutputFormat(AvroKeyOutputFormat.class).setEnableExploreOnCreate(true).setSerDe("org.apache.hadoop.hive.serde2.avro.AvroSerDe").setExploreInputFormat("org.apache.hadoop.hive.ql.io.avro.AvroContainerInputFormat").setExploreOutputFormat("org.apache.hadoop.hive.ql.io.avro.AvroContainerOutputFormat").setTableProperty("avro.schema.literal", Constants.ERROR_SCHEMA.toString()).build(), config.getEngine());
    BatchPipelineSpec spec = specGenerator.generateSpec(config);
    addWorkflow(new SmartWorkflow(spec, supportedPluginTypes, getConfigurer(), config.getEngine()));
    Schedules.Builder scheduleBuilder = Schedules.builder(SCHEDULE_NAME).setDescription("Data pipeline schedule");
    Integer maxConcurrentRuns = config.getMaxConcurrentRuns();
    if (maxConcurrentRuns != null) {
        scheduleBuilder.setMaxConcurrentRuns(maxConcurrentRuns);
    }
    scheduleWorkflow(scheduleBuilder.createTimeSchedule(config.getSchedule()), SmartWorkflow.NAME);
}

Example 44 with ETLBatchConfig

use of co.cask.cdap.etl.proto.v2.ETLBatchConfig in project cdap by caskdata.

the class DataPipelineTest method testInnerJoinWithMultiOutput.

public void testInnerJoinWithMultiOutput(Engine engine) throws Exception {
    Schema inputSchema1 = Schema.recordOf("customerRecord", Schema.Field.of("customer_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("customer_name", Schema.of(Schema.Type.STRING)));
    Schema inputSchema2 = Schema.recordOf("itemRecord", Schema.Field.of("item_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("item_price", Schema.of(Schema.Type.LONG)), Schema.Field.of("cust_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("cust_name", Schema.of(Schema.Type.STRING)));
    Schema inputSchema3 = Schema.recordOf("transactionRecord", Schema.Field.of("t_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("c_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("c_name", Schema.of(Schema.Type.STRING)));
    String input1Name = "source1InnerJoinInput-" + engine;
    String input2Name = "source2InnerJoinInput-" + engine;
    String input3Name = "source3InnerJoinInput-" + engine;
    String outputName = "innerJoinOutput-" + engine;
    String outputName2 = "innerJoinOutput2-" + engine;
    String joinerName = "innerJoiner-" + engine;
    String sinkName = "innerJoinSink-" + engine;
    String sinkName2 = "innerJoinSink-2" + engine;
    ETLBatchConfig etlConfig = ETLBatchConfig.builder("* * * * *").addStage(new ETLStage("source1", MockSource.getPlugin(input1Name, inputSchema1))).addStage(new ETLStage("source2", MockSource.getPlugin(input2Name, inputSchema2))).addStage(new ETLStage("source3", MockSource.getPlugin(input3Name, inputSchema3))).addStage(new ETLStage("t1", IdentityTransform.getPlugin())).addStage(new ETLStage("t2", IdentityTransform.getPlugin())).addStage(new ETLStage("t3", IdentityTransform.getPlugin())).addStage(new ETLStage(joinerName, MockJoiner.getPlugin("t1.customer_id=t2.cust_id=t3.c_id&" + "t1.customer_name=t2.cust_name=t3.c_name", "t1,t2,t3", ""))).addStage(new ETLStage(sinkName, MockSink.getPlugin(outputName))).addStage(new ETLStage(sinkName2, MockSink.getPlugin(outputName2))).addConnection("source1", "t1").addConnection("source2", "t2").addConnection("source3", "t3").addConnection("t1", joinerName).addConnection("t2", joinerName).addConnection("t3", joinerName).addConnection(joinerName, sinkName).addConnection(joinerName, sinkName2).setEngine(engine).build();
    AppRequest<ETLBatchConfig> appRequest = new AppRequest<>(APP_ARTIFACT, etlConfig);
    ApplicationId appId = NamespaceId.DEFAULT.app("InnerJoinApp-" + engine);
    ApplicationManager appManager = deployApplication(appId.toId(), appRequest);
    Schema outSchema = Schema.recordOf("join.output", Schema.Field.of("customer_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("customer_name", Schema.of(Schema.Type.STRING)), Schema.Field.of("item_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("item_price", Schema.of(Schema.Type.LONG)), Schema.Field.of("cust_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("cust_name", Schema.of(Schema.Type.STRING)), Schema.Field.of("t_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("c_id", Schema.of(Schema.Type.STRING)), Schema.Field.of("c_name", Schema.of(Schema.Type.STRING)));
    StructuredRecord recordSamuel = StructuredRecord.builder(inputSchema1).set("customer_id", "1").set("customer_name", "samuel").build();
    StructuredRecord recordBob = StructuredRecord.builder(inputSchema1).set("customer_id", "2").set("customer_name", "bob").build();
    StructuredRecord recordJane = StructuredRecord.builder(inputSchema1).set("customer_id", "3").set("customer_name", "jane").build();
    StructuredRecord recordCar = StructuredRecord.builder(inputSchema2).set("item_id", "11").set("item_price", 10000L).set("cust_id", "1").set("cust_name", "samuel").build();
    StructuredRecord recordBike = StructuredRecord.builder(inputSchema2).set("item_id", "22").set("item_price", 100L).set("cust_id", "3").set("cust_name", "jane").build();
    StructuredRecord recordTrasCar = StructuredRecord.builder(inputSchema3).set("t_id", "1").set("c_id", "1").set("c_name", "samuel").build();
    StructuredRecord recordTrasBike = StructuredRecord.builder(inputSchema3).set("t_id", "2").set("c_id", "3").set("c_name", "jane").build();
    // write one record to each source
    DataSetManager<Table> inputManager = getDataset(NamespaceId.DEFAULT.dataset(input1Name));
    MockSource.writeInput(inputManager, ImmutableList.of(recordSamuel, recordBob, recordJane));
    inputManager = getDataset(NamespaceId.DEFAULT.dataset(input2Name));
    MockSource.writeInput(inputManager, ImmutableList.of(recordCar, recordBike));
    inputManager = getDataset(NamespaceId.DEFAULT.dataset(input3Name));
    MockSource.writeInput(inputManager, ImmutableList.of(recordTrasCar, recordTrasBike));
    WorkflowManager workflowManager = appManager.getWorkflowManager(SmartWorkflow.NAME);
    workflowManager.start();
    workflowManager.waitForRun(ProgramRunStatus.COMPLETED, 5, TimeUnit.MINUTES);
    StructuredRecord joinRecordSamuel = StructuredRecord.builder(outSchema).set("customer_id", "1").set("customer_name", "samuel").set("item_id", "11").set("item_price", 10000L).set("cust_id", "1").set("cust_name", "samuel").set("t_id", "1").set("c_id", "1").set("c_name", "samuel").build();
    StructuredRecord joinRecordJane = StructuredRecord.builder(outSchema).set("customer_id", "3").set("customer_name", "jane").set("item_id", "22").set("item_price", 100L).set("cust_id", "3").set("cust_name", "jane").set("t_id", "2").set("c_id", "3").set("c_name", "jane").build();
    DataSetManager<Table> sinkManager = getDataset(outputName);
    Set<StructuredRecord> expected = ImmutableSet.of(joinRecordSamuel, joinRecordJane);
    Set<StructuredRecord> actual = Sets.newHashSet(MockSink.readOutput(sinkManager));
    Assert.assertEquals(expected, actual);
    sinkManager = getDataset(outputName2);
    actual = Sets.newHashSet(MockSink.readOutput(sinkManager));
    Assert.assertEquals(expected, actual);
    validateMetric(2, appId, joinerName + ".records.out");
    validateMetric(2, appId, sinkName + ".records.in");
    validateMetric(2, appId, sinkName2 + ".records.in");
}

Example 45 with ETLBatchConfig

use of co.cask.cdap.etl.proto.v2.ETLBatchConfig in project cdap by caskdata.

the class DataPipelineTest method testParallelAggregators.

private void testParallelAggregators(Engine engine) throws Exception {
    String source1Name = "pAggInput1-" + engine.name();
    String source2Name = "pAggInput2-" + engine.name();
    String sink1Name = "pAggOutput1-" + engine.name();
    String sink2Name = "pAggOutput2-" + engine.name();
    Schema inputSchema = Schema.recordOf("testRecord", Schema.Field.of("user", Schema.of(Schema.Type.STRING)), Schema.Field.of("item", Schema.of(Schema.Type.LONG)));
    /*
       source1 --|--> agg1 --> sink1
                 |
       source2 --|--> agg2 --> sink2
     */
    ETLBatchConfig etlConfig = ETLBatchConfig.builder("* * * * *").setEngine(engine).addStage(new ETLStage("source1", MockSource.getPlugin(source1Name, inputSchema))).addStage(new ETLStage("source2", MockSource.getPlugin(source2Name, inputSchema))).addStage(new ETLStage("sink1", MockSink.getPlugin(sink1Name))).addStage(new ETLStage("sink2", MockSink.getPlugin(sink2Name))).addStage(new ETLStage("agg1", FieldCountAggregator.getPlugin("user", "string"))).addStage(new ETLStage("agg2", FieldCountAggregator.getPlugin("item", "long"))).addConnection("source1", "agg1").addConnection("source1", "agg2").addConnection("source2", "agg1").addConnection("source2", "agg2").addConnection("agg1", "sink1").addConnection("agg2", "sink2").build();
    AppRequest<ETLBatchConfig> appRequest = new AppRequest<>(APP_ARTIFACT, etlConfig);
    ApplicationId appId = NamespaceId.DEFAULT.app("ParallelAggApp-" + engine);
    ApplicationManager appManager = deployApplication(appId.toId(), appRequest);
    // write few records to each source
    DataSetManager<Table> inputManager = getDataset(NamespaceId.DEFAULT.dataset(source1Name));
    MockSource.writeInput(inputManager, ImmutableList.of(StructuredRecord.builder(inputSchema).set("user", "samuel").set("item", 1L).build(), StructuredRecord.builder(inputSchema).set("user", "samuel").set("item", 2L).build()));
    inputManager = getDataset(NamespaceId.DEFAULT.dataset(source2Name));
    MockSource.writeInput(inputManager, ImmutableList.of(StructuredRecord.builder(inputSchema).set("user", "samuel").set("item", 3L).build(), StructuredRecord.builder(inputSchema).set("user", "john").set("item", 4L).build(), StructuredRecord.builder(inputSchema).set("user", "john").set("item", 3L).build()));
    WorkflowManager workflowManager = appManager.getWorkflowManager(SmartWorkflow.NAME);
    workflowManager.start();
    workflowManager.waitForRun(ProgramRunStatus.COMPLETED, 5, TimeUnit.MINUTES);
    Schema outputSchema1 = Schema.recordOf("user.count", Schema.Field.of("user", Schema.of(Schema.Type.STRING)), Schema.Field.of("ct", Schema.of(Schema.Type.LONG)));
    Schema outputSchema2 = Schema.recordOf("item.count", Schema.Field.of("item", Schema.of(Schema.Type.LONG)), Schema.Field.of("ct", Schema.of(Schema.Type.LONG)));
    // check output
    DataSetManager<Table> sinkManager = getDataset(sink1Name);
    Set<StructuredRecord> expected = ImmutableSet.of(StructuredRecord.builder(outputSchema1).set("user", "all").set("ct", 5L).build(), StructuredRecord.builder(outputSchema1).set("user", "samuel").set("ct", 3L).build(), StructuredRecord.builder(outputSchema1).set("user", "john").set("ct", 2L).build());
    Set<StructuredRecord> actual = Sets.newHashSet(MockSink.readOutput(sinkManager));
    Assert.assertEquals(expected, actual);
    sinkManager = getDataset(sink2Name);
    expected = ImmutableSet.of(StructuredRecord.builder(outputSchema2).set("item", 0L).set("ct", 5L).build(), StructuredRecord.builder(outputSchema2).set("item", 1L).set("ct", 1L).build(), StructuredRecord.builder(outputSchema2).set("item", 2L).set("ct", 1L).build(), StructuredRecord.builder(outputSchema2).set("item", 3L).set("ct", 2L).build(), StructuredRecord.builder(outputSchema2).set("item", 4L).set("ct", 1L).build());
    actual = Sets.newHashSet(MockSink.readOutput(sinkManager));
    Assert.assertEquals(expected, actual);
    validateMetric(2, appId, "source1.records.out");
    validateMetric(3, appId, "source2.records.out");
    validateMetric(5, appId, "agg1.records.in");
    // 2 users, but FieldCountAggregator always emits an 'all' group
    validateMetric(3, appId, "agg1.aggregator.groups");
    validateMetric(3, appId, "agg1.records.out");
    validateMetric(5, appId, "agg2.records.in");
    // 4 items, but FieldCountAggregator always emits an 'all' group
    validateMetric(5, appId, "agg2.aggregator.groups");
    validateMetric(5, appId, "agg2.records.out");
    validateMetric(3, appId, "sink1.records.in");
    validateMetric(5, appId, "sink2.records.in");
}