Examples with DataflowPackage com.google.api.services.dataflow.model.DataflowPackage used on opensource projects

Example 16 with DataflowPackage

use of com.google.api.services.dataflow.model.DataflowPackage in project beam by apache.

the class DataflowPipelineTranslatorTest method createPredefinedStep.

/**
   * Returns a Step for a {@link DoFn} by creating and translating a pipeline.
   */
private static Step createPredefinedStep() throws Exception {
    DataflowPipelineOptions options = buildPipelineOptions();
    DataflowPipelineTranslator translator = DataflowPipelineTranslator.fromOptions(options);
    Pipeline pipeline = Pipeline.create(options);
    String stepName = "DoFn1";
    pipeline.apply("ReadMyFile", TextIO.read().from("gs://bucket/in")).apply(stepName, ParDo.of(new NoOpFn())).apply("WriteMyFile", TextIO.write().to("gs://bucket/out"));
    DataflowRunner runner = DataflowRunner.fromOptions(options);
    runner.replaceTransforms(pipeline);
    Job job = translator.translate(pipeline, runner, Collections.<DataflowPackage>emptyList()).getJob();
    assertEquals(8, job.getSteps().size());
    Step step = job.getSteps().get(1);
    assertEquals(stepName, getString(step.getProperties(), PropertyNames.USER_NAME));
    assertAllStepOutputsHaveUniqueIds(job);
    return step;
}

Example 17 with DataflowPackage

use of com.google.api.services.dataflow.model.DataflowPackage in project beam by apache.

the class DataflowPipelineTranslatorTest method testSubnetworkConfig.

@Test
public void testSubnetworkConfig() throws IOException {
    final String testSubnetwork = "regions/REGION/subnetworks/SUBNETWORK";
    DataflowPipelineOptions options = buildPipelineOptions();
    options.setSubnetwork(testSubnetwork);
    Pipeline p = buildPipeline(options);
    p.traverseTopologically(new RecordingPipelineVisitor());
    Job job = DataflowPipelineTranslator.fromOptions(options).translate(p, DataflowRunner.fromOptions(options), Collections.<DataflowPackage>emptyList()).getJob();
    assertEquals(1, job.getEnvironment().getWorkerPools().size());
    assertEquals(testSubnetwork, job.getEnvironment().getWorkerPools().get(0).getSubnetwork());
}

Example 18 with DataflowPackage

use of com.google.api.services.dataflow.model.DataflowPackage in project beam by apache.

the class DataflowPipelineTranslatorTest method testSettingOfPipelineOptionsWithCustomUserType.

@Test
public void testSettingOfPipelineOptionsWithCustomUserType() throws IOException {
    DataflowPipelineOptions options = buildPipelineOptions();
    options.setRunner(DataflowRunner.class);
    options.as(JacksonIncompatibleOptions.class).setJacksonIncompatible(new JacksonIncompatible("userCustomTypeTest"));
    Pipeline p = Pipeline.create(options);
    p.traverseTopologically(new RecordingPipelineVisitor());
    Job job = DataflowPipelineTranslator.fromOptions(options).translate(p, DataflowRunner.fromOptions(options), Collections.<DataflowPackage>emptyList()).getJob();
    Map<String, Object> sdkPipelineOptions = job.getEnvironment().getSdkPipelineOptions();
    assertThat(sdkPipelineOptions, hasKey("options"));
    Map<String, Object> optionsMap = (Map<String, Object>) sdkPipelineOptions.get("options");
    assertThat(optionsMap, hasEntry("jacksonIncompatible", (Object) "userCustomTypeTest"));
}

Example 19 with DataflowPackage

use of com.google.api.services.dataflow.model.DataflowPackage in project beam by apache.

the class DataflowPipelineTranslatorTest method testPredefinedAddStep.

@Test
public void testPredefinedAddStep() throws Exception {
    DataflowPipelineOptions options = buildPipelineOptions();
    DataflowPipelineTranslator translator = DataflowPipelineTranslator.fromOptions(options);
    DataflowPipelineTranslator.registerTransformTranslator(EmbeddedTransform.class, new EmbeddedTranslator());
    // Create a predefined step using another pipeline
    Step predefinedStep = createPredefinedStep();
    // Create a pipeline that the predefined step will be embedded into
    Pipeline pipeline = Pipeline.create(options);
    pipeline.apply("ReadMyFile", TextIO.read().from("gs://bucket/in")).apply(ParDo.of(new NoOpFn())).apply(new EmbeddedTransform(predefinedStep.clone())).apply(ParDo.of(new NoOpFn()));
    DataflowRunner runner = DataflowRunner.fromOptions(options);
    runner.replaceTransforms(pipeline);
    Job job = translator.translate(pipeline, runner, Collections.<DataflowPackage>emptyList()).getJob();
    assertAllStepOutputsHaveUniqueIds(job);
    List<Step> steps = job.getSteps();
    assertEquals(4, steps.size());
    // The input to the embedded step should match the output of the step before
    Map<String, Object> step1Out = getOutputPortReference(steps.get(1));
    Map<String, Object> step2In = getDictionary(steps.get(2).getProperties(), PropertyNames.PARALLEL_INPUT);
    assertEquals(step1Out, step2In);
    // The output from the embedded step should match the input of the step after
    Map<String, Object> step2Out = getOutputPortReference(steps.get(2));
    Map<String, Object> step3In = getDictionary(steps.get(3).getProperties(), PropertyNames.PARALLEL_INPUT);
    assertEquals(step2Out, step3In);
    // The step should not have been modified other than remapping the input
    Step predefinedStepClone = predefinedStep.clone();
    Step embeddedStepClone = steps.get(2).clone();
    predefinedStepClone.getProperties().remove(PropertyNames.PARALLEL_INPUT);
    embeddedStepClone.getProperties().remove(PropertyNames.PARALLEL_INPUT);
    assertEquals(predefinedStepClone, embeddedStepClone);
}

Example 20 with DataflowPackage

use of com.google.api.services.dataflow.model.DataflowPackage in project beam by apache.

the class DataflowPipelineTranslatorTest method testSettingOfSdkPipelineOptions.

@Test
public void testSettingOfSdkPipelineOptions() throws IOException {
    DataflowPipelineOptions options = buildPipelineOptions();
    options.setRunner(DataflowRunner.class);
    Pipeline p = Pipeline.create(options);
    p.traverseTopologically(new RecordingPipelineVisitor());
    Job job = DataflowPipelineTranslator.fromOptions(options).translate(p, DataflowRunner.fromOptions(options), Collections.<DataflowPackage>emptyList()).getJob();
    Map<String, Object> sdkPipelineOptions = job.getEnvironment().getSdkPipelineOptions();
    assertThat(sdkPipelineOptions, hasKey("options"));
    Map<String, Object> optionsMap = (Map<String, Object>) sdkPipelineOptions.get("options");
    assertThat(optionsMap, hasEntry("appName", (Object) "DataflowPipelineTranslatorTest"));
    assertThat(optionsMap, hasEntry("project", (Object) "some-project"));
    assertThat(optionsMap, hasEntry("pathValidatorClass", (Object) GcsPathValidator.class.getName()));
    assertThat(optionsMap, hasEntry("runner", (Object) DataflowRunner.class.getName()));
    assertThat(optionsMap, hasEntry("jobName", (Object) "some-job-name"));
    assertThat(optionsMap, hasEntry("tempLocation", (Object) "gs://somebucket/some/path"));
    assertThat(optionsMap, hasEntry("stagingLocation", (Object) "gs://somebucket/some/path/staging/"));
    assertThat(optionsMap, hasEntry("stableUniqueNames", (Object) "WARNING"));
    assertThat(optionsMap, hasEntry("streaming", (Object) false));
    assertThat(optionsMap, hasEntry("numberOfWorkerHarnessThreads", (Object) 0));
}