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

Example 6 with ETLBatchConfig

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

the class DataPipelineTest method testErrorTransform.

private void testErrorTransform(Engine engine) throws Exception {
    String source1TableName = "errTestIn1-" + engine;
    String source2TableName = "errTestIn2-" + engine;
    String sink1TableName = "errTestOut1-" + engine;
    String sink2TableName = "errTestOut2-" + engine;
    Schema inputSchema = Schema.recordOf("user", Schema.Field.of("name", Schema.of(Schema.Type.STRING)), Schema.Field.of("id", Schema.of(Schema.Type.INT)));
    /*
     *
     * source1 --> filter1 --> filter2 --> agg1 --> agg2
     *                |           |         |        |
     *                |-----------|---------|--------|--------|--> errorflatten --> sink1
     *                |                                       |
     *                |                                       |--> errorfilter --> sink2
     *                |
     * source2 --> dropnull
     *
     * arrows coming out the right represent output records
     * arrows coming out the bottom represent error records
     * this will test multiple stages from multiple phases emitting errors to the same stage
     * as well as errors from one stage going to multiple stages
     * and transforms that have an error schema different from their output schema
     */
    ETLBatchConfig config = ETLBatchConfig.builder("* * * * *").setEngine(engine).addStage(new ETLStage("source1", MockSource.getPlugin(source1TableName, inputSchema))).addStage(new ETLStage("source2", MockSource.getPlugin(source2TableName, inputSchema))).addStage(new ETLStage("filter1", StringValueFilterTransform.getPlugin("name", "Leo"))).addStage(new ETLStage("filter2", StringValueFilterTransform.getPlugin("name", "Ralph"))).addStage(new ETLStage("agg1", GroupFilterAggregator.getPlugin("name", "Don"))).addStage(new ETLStage("agg2", GroupFilterAggregator.getPlugin("name", "Mike"))).addStage(new ETLStage("errorflatten", FlattenErrorTransform.getPlugin())).addStage(new ETLStage("errorfilter", FilterErrorTransform.getPlugin(3))).addStage(new ETLStage("dropnull", DropNullTransform.getPlugin("name"))).addStage(new ETLStage("sink1", MockSink.getPlugin(sink1TableName))).addStage(new ETLStage("sink2", MockSink.getPlugin(sink2TableName))).addConnection("source1", "filter1").addConnection("source2", "dropnull").addConnection("filter1", "filter2").addConnection("filter2", "agg1").addConnection("agg1", "agg2").addConnection("filter1", "errorflatten").addConnection("filter1", "errorfilter").addConnection("filter2", "errorflatten").addConnection("filter2", "errorfilter").addConnection("agg1", "errorflatten").addConnection("agg1", "errorfilter").addConnection("agg2", "errorflatten").addConnection("agg2", "errorfilter").addConnection("dropnull", "errorflatten").addConnection("dropnull", "errorfilter").addConnection("errorflatten", "sink1").addConnection("errorfilter", "sink2").build();
    AppRequest<ETLBatchConfig> appRequest = new AppRequest<>(APP_ARTIFACT, config);
    ApplicationId appId = NamespaceId.DEFAULT.app("ErrTransformTest-" + engine);
    ApplicationManager appManager = deployApplication(appId, appRequest);
    List<StructuredRecord> input = ImmutableList.of(StructuredRecord.builder(inputSchema).set("name", "Leo").set("id", 1).build(), StructuredRecord.builder(inputSchema).set("name", "Ralph").set("id", 2).build(), StructuredRecord.builder(inputSchema).set("name", "Don").set("id", 3).build(), StructuredRecord.builder(inputSchema).set("name", "Mike").set("id", 4).build());
    DataSetManager<Table> source1Table = getDataset(source1TableName);
    MockSource.writeInput(source1Table, input);
    input = ImmutableList.of(StructuredRecord.builder(inputSchema).set("name", "April").set("id", 5).build());
    DataSetManager<Table> source2Table = getDataset(source2TableName);
    MockSource.writeInput(source2Table, input);
    WorkflowManager workflowManager = appManager.getWorkflowManager(SmartWorkflow.NAME);
    workflowManager.start();
    workflowManager.waitForRun(ProgramRunStatus.COMPLETED, 5, TimeUnit.MINUTES);
    Schema flattenSchema = Schema.recordOf("erroruser", Schema.Field.of("name", Schema.of(Schema.Type.STRING)), Schema.Field.of("id", Schema.of(Schema.Type.INT)), Schema.Field.of("errMsg", Schema.nullableOf(Schema.of(Schema.Type.STRING))), Schema.Field.of("errCode", Schema.nullableOf(Schema.of(Schema.Type.INT))), Schema.Field.of("errStage", Schema.nullableOf(Schema.of(Schema.Type.STRING))));
    Set<StructuredRecord> expected = ImmutableSet.of(StructuredRecord.builder(flattenSchema).set("name", "Leo").set("id", 1).set("errMsg", "bad string value").set("errCode", 1).set("errStage", "filter1").build(), StructuredRecord.builder(flattenSchema).set("name", "Ralph").set("id", 2).set("errMsg", "bad string value").set("errCode", 1).set("errStage", "filter2").build(), StructuredRecord.builder(flattenSchema).set("name", "Don").set("id", 3).set("errMsg", "bad val").set("errCode", 3).set("errStage", "agg1").build(), StructuredRecord.builder(flattenSchema).set("name", "Mike").set("id", 4).set("errMsg", "bad val").set("errCode", 3).set("errStage", "agg2").build(), StructuredRecord.builder(flattenSchema).set("name", "April").set("id", 5).set("errMsg", "Field name was not null").set("errCode", 5).set("errStage", "dropnull").build());
    DataSetManager<Table> sink1Table = getDataset(sink1TableName);
    Assert.assertEquals(expected, ImmutableSet.copyOf(MockSink.readOutput(sink1Table)));
    expected = ImmutableSet.of(StructuredRecord.builder(inputSchema).set("name", "Leo").set("id", 1).build(), StructuredRecord.builder(inputSchema).set("name", "Ralph").set("id", 2).build(), StructuredRecord.builder(inputSchema).set("name", "April").set("id", 5).build());
    DataSetManager<Table> sink2Table = getDataset(sink2TableName);
    Assert.assertEquals(expected, ImmutableSet.copyOf(MockSink.readOutput(sink2Table)));
    /*
     *
     * source1 (4) --> filter1 (3) --> filter2 (2) --> agg1 (1) --> agg2
     *                   |                |              |            |
     *                  (1)              (1)            (1)          (1)
     *                   |----------------|--------------|------------|--------|--> errorflatten (5) --> sink1
     *                   |                                                     |
     *                  (1)                                                    |--> errorfilter (3) --> sink2
     *                   |
     * source2 --> dropnull
     */
    validateMetric(4, appId, "source1.records.out");
    validateMetric(1, appId, "source2.records.out");
    validateMetric(1, appId, "dropnull.records.error");
    validateMetric(3, appId, "filter1.records.out");
    validateMetric(1, appId, "filter1.records.error");
    validateMetric(2, appId, "filter2.records.out");
    validateMetric(1, appId, "filter2.records.error");
    validateMetric(1, appId, "agg1.records.out");
    validateMetric(1, appId, "agg1.records.error");
    validateMetric(1, appId, "agg2.records.error");
    validateMetric(5, appId, "errorflatten.records.out");
    validateMetric(3, appId, "errorfilter.records.out");
    validateMetric(5, appId, "sink1.records.in");
    validateMetric(3, appId, "sink2.records.in");
}

Example 7 with ETLBatchConfig

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

the class DataPipelineTest method testSimpleConditionWithMultipleInputActions.

@Test
public void testSimpleConditionWithMultipleInputActions() throws Exception {
    Schema schema = Schema.recordOf("testRecord", Schema.Field.of("name", Schema.of(Schema.Type.STRING)));
    /*
     * action1--|
     *          |--> condition --> file ---> trueSink
     * action2--|      |
     *                 |--->file----> falseSink
     *
     */
    String appName = "SimpleConditionWithMultipleInputActions";
    String trueSource = "true" + appName + "Source";
    String falseSource = "false" + appName + "Source";
    String trueSink = "true" + appName + "Sink";
    String falseSink = "false" + appName + "Sink";
    String actionTable = "actionTable" + appName;
    ETLBatchConfig etlConfig = ETLBatchConfig.builder("* * * * *").addStage(new ETLStage("trueSource", MockSource.getPlugin(trueSource, schema))).addStage(new ETLStage("falseSource", MockSource.getPlugin(falseSource, schema))).addStage(new ETLStage("trueSink", MockSink.getPlugin(trueSink))).addStage(new ETLStage("falseSink", MockSink.getPlugin(falseSink))).addStage(new ETLStage("condition", MockCondition.getPlugin("condition"))).addStage(new ETLStage("action1", MockAction.getPlugin(actionTable, "row1", "key1", "val1"))).addStage(new ETLStage("action2", MockAction.getPlugin(actionTable, "row2", "key2", "val2"))).addConnection("action1", "condition").addConnection("action2", "condition").addConnection("condition", "trueSource", true).addConnection("condition", "falseSource", false).addConnection("trueSource", "trueSink").addConnection("falseSource", "falseSink").build();
    AppRequest<ETLBatchConfig> appRequest = new AppRequest<>(APP_ARTIFACT_RANGE, etlConfig);
    ApplicationId appId = NamespaceId.DEFAULT.app(appName);
    ApplicationManager appManager = deployApplication(appId, appRequest);
    StructuredRecord recordSamuel = StructuredRecord.builder(schema).set("name", "samuel").build();
    StructuredRecord recordBob = StructuredRecord.builder(schema).set("name", "bob").build();
    for (String branch : Arrays.asList("true", "false")) {
        // write records to source
        String source = branch.equals("true") ? trueSource : falseSource;
        String sink = branch.equals("true") ? trueSink : falseSink;
        DataSetManager<Table> inputManager = getDataset(NamespaceId.DEFAULT.dataset(source));
        MockSource.writeInput(inputManager, ImmutableList.of(recordSamuel, recordBob));
        WorkflowManager workflowManager = appManager.getWorkflowManager(SmartWorkflow.NAME);
        workflowManager.start(ImmutableMap.of("condition.branch.to.execute", branch));
        if (branch.equals("true")) {
            workflowManager.waitForRun(ProgramRunStatus.COMPLETED, 5, TimeUnit.MINUTES);
        } else {
            workflowManager.waitForRuns(ProgramRunStatus.COMPLETED, 2, 5, TimeUnit.MINUTES);
        }
        // check sink
        DataSetManager<Table> sinkManager = getDataset(sink);
        Set<StructuredRecord> expected = ImmutableSet.of(recordSamuel, recordBob);
        Set<StructuredRecord> actual = Sets.newHashSet(MockSink.readOutput(sinkManager));
        Assert.assertEquals(expected, actual);
        validateMetric(2, appId, branch + "Source.records.out");
        validateMetric(2, appId, branch + "Sink.records.in");
        // check Action1 and Action2 is executed correctly
        DataSetManager<Table> actionTableDS = getDataset(actionTable);
        Assert.assertEquals("val1", MockAction.readOutput(actionTableDS, "row1", "key1"));
        Assert.assertEquals("val2", MockAction.readOutput(actionTableDS, "row2", "key2"));
    }
}

Example 8 with ETLBatchConfig

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

the class DataPipelineTest method testMultipleOrderedInputActions.

@Test
public void testMultipleOrderedInputActions() throws Exception {
    Schema schema = Schema.recordOf("testRecord", Schema.Field.of("name", Schema.of(Schema.Type.STRING)));
    /*
     * action1--->action2---|
     *                      |--> condition --> file ---> trueSink
     * action3--->action4---|      |
     *                             |--->file----> falseSink
     *
     */
    String appName = "MultipleOrderedInputActions";
    String trueSource = "true" + appName + "Source";
    String falseSource = "false" + appName + "Source";
    String trueSink = "true" + appName + "Sink";
    String falseSink = "false" + appName + "Sink";
    String actionTable = "actionTable" + appName;
    ETLBatchConfig etlConfig = ETLBatchConfig.builder("* * * * *").addStage(new ETLStage("trueSource", MockSource.getPlugin(trueSource, schema))).addStage(new ETLStage("falseSource", MockSource.getPlugin(falseSource, schema))).addStage(new ETLStage("trueSink", MockSink.getPlugin(trueSink))).addStage(new ETLStage("falseSink", MockSink.getPlugin(falseSink))).addStage(new ETLStage("condition", MockCondition.getPlugin("condition"))).addStage(new ETLStage("action1", MockAction.getPlugin(actionTable, "row1", "key1", "val1"))).addStage(new ETLStage("action2", MockAction.getPlugin(actionTable, "row2", "key2", "val2", "row1key1", "val1"))).addStage(new ETLStage("action3", MockAction.getPlugin(actionTable, "row3", "key3", "val3"))).addStage(new ETLStage("action4", MockAction.getPlugin(actionTable, "row4", "key4", "val4", "row3key3", "val3"))).addConnection("action1", "action2").addConnection("action3", "action4").addConnection("action2", "condition").addConnection("action4", "condition").addConnection("condition", "trueSource", true).addConnection("condition", "falseSource", false).addConnection("trueSource", "trueSink").addConnection("falseSource", "falseSink").build();
    AppRequest<ETLBatchConfig> appRequest = new AppRequest<>(APP_ARTIFACT_RANGE, etlConfig);
    ApplicationId appId = NamespaceId.DEFAULT.app(appName);
    ApplicationManager appManager = deployApplication(appId, appRequest);
    StructuredRecord recordSamuel = StructuredRecord.builder(schema).set("name", "samuel").build();
    StructuredRecord recordBob = StructuredRecord.builder(schema).set("name", "bob").build();
    for (String branch : Arrays.asList("true", "false")) {
        // write records to source
        String source = branch.equals("true") ? trueSource : falseSource;
        String sink = branch.equals("true") ? trueSink : falseSink;
        DataSetManager<Table> inputManager = getDataset(NamespaceId.DEFAULT.dataset(source));
        MockSource.writeInput(inputManager, ImmutableList.of(recordSamuel, recordBob));
        WorkflowManager workflowManager = appManager.getWorkflowManager(SmartWorkflow.NAME);
        workflowManager.start(ImmutableMap.of("condition.branch.to.execute", branch));
        if (branch.equals("true")) {
            workflowManager.waitForRun(ProgramRunStatus.COMPLETED, 5, TimeUnit.MINUTES);
        } else {
            workflowManager.waitForRuns(ProgramRunStatus.COMPLETED, 2, 5, TimeUnit.MINUTES);
        }
        // check sink
        DataSetManager<Table> sinkManager = getDataset(sink);
        Set<StructuredRecord> expected = ImmutableSet.of(recordSamuel, recordBob);
        Set<StructuredRecord> actual = Sets.newHashSet(MockSink.readOutput(sinkManager));
        Assert.assertEquals(expected, actual);
        validateMetric(2, appId, branch + "Source.records.out");
        validateMetric(2, appId, branch + "Sink.records.in");
        // check Action1 and Action2 is executed correctly
        DataSetManager<Table> actionTableDS = getDataset(actionTable);
        Assert.assertEquals("val1", MockAction.readOutput(actionTableDS, "row1", "key1"));
        Assert.assertEquals("val2", MockAction.readOutput(actionTableDS, "row2", "key2"));
        Assert.assertEquals("val3", MockAction.readOutput(actionTableDS, "row3", "key3"));
        Assert.assertEquals("val4", MockAction.readOutput(actionTableDS, "row4", "key4"));
    }
}

Example 9 with ETLBatchConfig

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

the class DataPipelineTest method testPipelineWithActions.

private void testPipelineWithActions(Engine engine) throws Exception {
    String actionTable = "actionTable-" + engine;
    String action1RowKey = "action1.row";
    String action1ColumnKey = "action1.column";
    String action1Value = "action1.value";
    String action2RowKey = "action2.row";
    String action2ColumnKey = "action2.column";
    String action2Value = "action2.value";
    String action3RowKey = "action3.row";
    String action3ColumnKey = "action3.column";
    String action3Value = "action3.value";
    String sourceName = "actionSource-" + engine;
    String sinkName = "actionSink-" + engine;
    String sourceTableName = "actionSourceTable-" + engine;
    String sinkTableName = "actionSinkTable-" + engine;
    Schema schema = Schema.recordOf("testRecord", Schema.Field.of("name", Schema.of(Schema.Type.STRING)));
    ETLBatchConfig etlConfig = ETLBatchConfig.builder("* * * * *").addStage(new ETLStage("action1", MockAction.getPlugin(actionTable, action1RowKey, action1ColumnKey, action1Value))).addStage(new ETLStage("action2", MockAction.getPlugin(actionTable, action2RowKey, action2ColumnKey, action2Value))).addStage(new ETLStage("action3", MockAction.getPlugin(actionTable, action3RowKey, action3ColumnKey, action3Value))).addStage(new ETLStage(sourceName, MockSource.getPlugin(sourceTableName, schema))).addStage(new ETLStage(sinkName, MockSink.getPlugin(sinkTableName))).addConnection(sourceName, sinkName).addConnection("action1", "action2").addConnection("action2", sourceName).addConnection(sinkName, "action3").setEngine(engine).build();
    AppRequest<ETLBatchConfig> appRequest = new AppRequest<>(APP_ARTIFACT, etlConfig);
    ApplicationId appId = NamespaceId.DEFAULT.app("MyApp-" + engine);
    ApplicationManager appManager = deployApplication(appId, appRequest);
    StructuredRecord recordSamuel = StructuredRecord.builder(schema).set("name", "samuel").build();
    StructuredRecord recordBob = StructuredRecord.builder(schema).set("name", "bob").build();
    // write records to source
    DataSetManager<Table> inputManager = getDataset(NamespaceId.DEFAULT.dataset(sourceTableName));
    MockSource.writeInput(inputManager, ImmutableList.of(recordSamuel, recordBob));
    WorkflowManager workflowManager = appManager.getWorkflowManager(SmartWorkflow.NAME);
    workflowManager.start();
    workflowManager.waitForRun(ProgramRunStatus.COMPLETED, 5, TimeUnit.MINUTES);
    // check sink
    DataSetManager<Table> sinkManager = getDataset(sinkTableName);
    Set<StructuredRecord> expected = ImmutableSet.of(recordSamuel, recordBob);
    Set<StructuredRecord> actual = Sets.newHashSet(MockSink.readOutput(sinkManager));
    Assert.assertEquals(expected, actual);
    DataSetManager<Table> actionTableDS = getDataset(actionTable);
    Assert.assertEquals(action1Value, MockAction.readOutput(actionTableDS, action1RowKey, action1ColumnKey));
    Assert.assertEquals(action2Value, MockAction.readOutput(actionTableDS, action2RowKey, action2ColumnKey));
    Assert.assertEquals(action3Value, MockAction.readOutput(actionTableDS, action3RowKey, action3ColumnKey));
    validateMetric(2, appId, sourceName + ".records.out");
    validateMetric(2, appId, sinkName + ".records.in");
    List<RunRecord> history = workflowManager.getHistory(ProgramRunStatus.COMPLETED);
    Assert.assertEquals(1, history.size());
    String runId = history.get(0).getPid();
    for (WorkflowToken.Scope scope : Arrays.asList(WorkflowToken.Scope.SYSTEM, WorkflowToken.Scope.USER)) {
        WorkflowTokenDetail token = workflowManager.getToken(runId, scope, null);
        for (Map.Entry<String, List<WorkflowTokenDetail.NodeValueDetail>> tokenData : token.getTokenData().entrySet()) {
            Assert.assertTrue(!tokenData.getKey().startsWith(co.cask.cdap.etl.common.Constants.StageStatistics.PREFIX));
        }
    }
}

Example 10 with ETLBatchConfig

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

the class DataPipelineTest method testRuntimeArgs.

private void testRuntimeArgs(Engine engine) throws Exception {
    String sourceName = "runtimeArgInput-" + engine;
    String sinkName = "runtimeArgOutput-" + engine;
    ETLBatchConfig etlConfig = ETLBatchConfig.builder("* * * * *").addStage(new ETLStage("action", MockAction.getPlugin("dumy", "val", "ue", "dwayne"))).addStage(new ETLStage("source", MockSource.getPlugin(sourceName))).addStage(new ETLStage("filter", StringValueFilterTransform.getPlugin("name", "samuel"))).addStage(new ETLStage("sink", MockSink.getPlugin(sinkName))).addConnection("action", "source").addConnection("source", "filter").addConnection("filter", "sink").setEngine(engine).build();
    AppRequest<ETLBatchConfig> appRequest = new AppRequest<>(APP_ARTIFACT, etlConfig);
    ApplicationId appId = NamespaceId.DEFAULT.app("RuntimeArgApp-" + engine);
    ApplicationManager appManager = deployApplication(appId, appRequest);
    // there should be only two programs - one workflow and one mapreduce/spark
    Schema schema = Schema.recordOf("testRecord", Schema.Field.of("name", Schema.of(Schema.Type.STRING)));
    StructuredRecord recordSamuel = StructuredRecord.builder(schema).set("name", "samuel").build();
    StructuredRecord recordDwayne = StructuredRecord.builder(schema).set("name", "dwayne").build();
    // write one record to each source
    DataSetManager<Table> inputManager = getDataset(sourceName);
    MockSource.writeInput(inputManager, ImmutableList.of(recordSamuel, recordDwayne));
    WorkflowManager workflowManager = appManager.getWorkflowManager(SmartWorkflow.NAME);
    workflowManager.start();
    workflowManager.waitForRun(ProgramRunStatus.COMPLETED, 5, TimeUnit.MINUTES);
    // check sink
    DataSetManager<Table> sinkManager = getDataset(sinkName);
    Set<StructuredRecord> expected = ImmutableSet.of(recordSamuel);
    Set<StructuredRecord> actual = Sets.newHashSet(MockSink.readOutput(sinkManager));
    Assert.assertEquals(expected, actual);
}