Examples with Pipeline org.apache.beam.model.pipeline.v1.RunnerApi.Pipeline used on opensource projects

Example 81 with Pipeline

use of org.apache.beam.model.pipeline.v1.RunnerApi.Pipeline in project beam by apache.

the class PortableTimersExecutionTest method testTimerExecution.

@Test(timeout = 120_000)
public void testTimerExecution() throws Exception {
    FlinkPipelineOptions options = PipelineOptionsFactory.fromArgs("--experiments=beam_fn_api").as(FlinkPipelineOptions.class);
    options.setRunner(CrashingRunner.class);
    options.setFlinkMaster("[local]");
    options.setStreaming(isStreaming);
    options.setParallelism(2);
    options.as(PortablePipelineOptions.class).setDefaultEnvironmentType(Environments.ENVIRONMENT_EMBEDDED);
    final String timerId = "foo";
    final String stateId = "sizzle";
    final int offset = 5000;
    final int timerOutput = 4093;
    // Enough keys that we exercise interesting code paths
    int numKeys = 50;
    int numDuplicateTimers = 15;
    List<KV<String, Integer>> input = new ArrayList<>();
    List<KV<String, Integer>> expectedOutput = new ArrayList<>();
    for (Integer key = 0; key < numKeys; ++key) {
        // Each key should have just one final output at GC time
        expectedOutput.add(KV.of(key.toString(), timerOutput));
        for (int i = 0; i < numDuplicateTimers; ++i) {
            // Each input should be output with the offset added
            input.add(KV.of(key.toString(), i));
            expectedOutput.add(KV.of(key.toString(), i + offset));
        }
    }
    Collections.shuffle(input);
    DoFn<byte[], KV<String, Integer>> inputFn = new DoFn<byte[], KV<String, Integer>>() {

        @ProcessElement
        public void processElement(ProcessContext context) {
            for (KV<String, Integer> stringIntegerKV : input) {
                context.output(stringIntegerKV);
            }
        }
    };
    DoFn<KV<String, Integer>, KV<String, Integer>> testFn = new DoFn<KV<String, Integer>, KV<String, Integer>>() {

        @TimerId(timerId)
        private final TimerSpec spec = TimerSpecs.timer(TimeDomain.EVENT_TIME);

        @StateId(stateId)
        private final StateSpec<ValueState<String>> stateSpec = StateSpecs.value(StringUtf8Coder.of());

        @ProcessElement
        public void processElement(ProcessContext context, @TimerId(timerId) Timer timer, @StateId(stateId) ValueState<String> state, BoundedWindow window) {
            timer.set(window.maxTimestamp());
            state.write(context.element().getKey());
            context.output(KV.of(context.element().getKey(), context.element().getValue() + offset));
        }

        @OnTimer(timerId)
        public void onTimer(@StateId(stateId) ValueState<String> state, OutputReceiver<KV<String, Integer>> r) {
            String read = Objects.requireNonNull(state.read(), "State must not be null");
            KV<String, Integer> of = KV.of(read, timerOutput);
            r.output(of);
        }
    };
    final Pipeline pipeline = Pipeline.create(options);
    PCollection<KV<String, Integer>> output = pipeline.apply("Impulse", Impulse.create()).apply("Input", ParDo.of(inputFn)).apply("Timers", ParDo.of(testFn));
    PAssert.that(output).containsInAnyOrder(expectedOutput);
    RunnerApi.Pipeline pipelineProto = PipelineTranslation.toProto(pipeline);
    JobInvocation jobInvocation = FlinkJobInvoker.create(null).createJobInvocation("id", "none", flinkJobExecutor, pipelineProto, options, new FlinkPipelineRunner(options, null, Collections.emptyList()));
    jobInvocation.start();
    while (jobInvocation.getState() != JobState.Enum.DONE) {
        Thread.sleep(1000);
    }
    assertThat(jobInvocation.getState(), is(JobState.Enum.DONE));
}

Example 82 with Pipeline

use of org.apache.beam.model.pipeline.v1.RunnerApi.Pipeline in project beam by apache.

the class DataflowPipelineTranslatorTest method testToMap.

@Test
public void testToMap() throws Exception {
    DataflowPipelineOptions options = buildPipelineOptions();
    Pipeline pipeline = Pipeline.create(options);
    final PCollectionView<Map<String, Integer>> view = pipeline.apply("CreateSideInput", Create.of(KV.of("a", 1), KV.of("b", 3))).apply(View.asMap());
    PCollection<KV<String, Integer>> output = pipeline.apply("CreateMainInput", Create.of("apple", "banana", "blackberry")).apply("OutputSideInputs", ParDo.of(new DoFn<String, KV<String, Integer>>() {

        @ProcessElement
        public void processElement(ProcessContext c) {
            c.output(KV.of(c.element(), c.sideInput(view).get(c.element().substring(0, 1))));
        }
    }).withSideInputs(view));
    PAssert.that(output).containsInAnyOrder(KV.of("apple", 1), KV.of("banana", 3), KV.of("blackberry", 3));
    DataflowRunner runner = DataflowRunner.fromOptions(options);
    DataflowPipelineTranslator translator = DataflowPipelineTranslator.fromOptions(options);
    runner.replaceV1Transforms(pipeline);
    SdkComponents sdkComponents = createSdkComponents(options);
    RunnerApi.Pipeline pipelineProto = PipelineTranslation.toProto(pipeline, sdkComponents, true);
    Job job = translator.translate(pipeline, pipelineProto, sdkComponents, runner, Collections.emptyList()).getJob();
    List<Step> steps = job.getSteps();
    // Change detector assertion just to make sure the test was not a noop.
    // No need to actually check the pipeline as the ValidatesRunner tests
    // ensure translation is correct. This is just a quick check to see that translation
    // does not crash.
    assertEquals(24, steps.size());
}

Example 83 with Pipeline

use of org.apache.beam.model.pipeline.v1.RunnerApi.Pipeline in project beam by apache.

the class DataflowPipelineTranslatorTest method testNumWorkersCannotExceedMaxNumWorkers.

@Test
public void testNumWorkersCannotExceedMaxNumWorkers() throws IOException {
    DataflowPipelineOptions options = buildPipelineOptions();
    options.setNumWorkers(43);
    options.setMaxNumWorkers(42);
    Pipeline p = buildPipeline(options);
    p.traverseTopologically(new RecordingPipelineVisitor());
    SdkComponents sdkComponents = createSdkComponents(options);
    RunnerApi.Pipeline pipelineProto = PipelineTranslation.toProto(p, sdkComponents, true);
    thrown.expect(IllegalArgumentException.class);
    thrown.expectMessage("numWorkers (43) cannot exceed maxNumWorkers (42).");
    DataflowPipelineTranslator.fromOptions(options).translate(p, pipelineProto, sdkComponents, DataflowRunner.fromOptions(options), Collections.emptyList()).getJob();
}

Example 84 with Pipeline

use of org.apache.beam.model.pipeline.v1.RunnerApi.Pipeline in project beam by apache.

the class DataflowPipelineTranslatorTest method testStepDisplayData.

@Test
public void testStepDisplayData() throws Exception {
    DataflowPipelineOptions options = buildPipelineOptions();
    DataflowPipelineTranslator translator = DataflowPipelineTranslator.fromOptions(options);
    Pipeline pipeline = Pipeline.create(options);
    DoFn<Integer, Integer> fn1 = new DoFn<Integer, Integer>() {

        @ProcessElement
        public void processElement(ProcessContext c) throws Exception {
            c.output(c.element());
        }

        @Override
        public void populateDisplayData(DisplayData.Builder builder) {
            builder.add(DisplayData.item("foo", "bar")).add(DisplayData.item("foo2", DataflowPipelineTranslatorTest.class).withLabel("Test Class").withLinkUrl("http://www.google.com"));
        }
    };
    DoFn<Integer, Integer> fn2 = new DoFn<Integer, Integer>() {

        @ProcessElement
        public void processElement(ProcessContext c) throws Exception {
            c.output(c.element());
        }

        @Override
        public void populateDisplayData(DisplayData.Builder builder) {
            builder.add(DisplayData.item("foo3", 1234));
        }
    };
    ParDo.SingleOutput<Integer, Integer> parDo1 = ParDo.of(fn1);
    ParDo.SingleOutput<Integer, Integer> parDo2 = ParDo.of(fn2);
    pipeline.apply(Create.of(1, 2, 3)).apply(parDo1).apply(parDo2);
    DataflowRunner runner = DataflowRunner.fromOptions(options);
    runner.replaceV1Transforms(pipeline);
    SdkComponents sdkComponents = createSdkComponents(options);
    RunnerApi.Pipeline pipelineProto = PipelineTranslation.toProto(pipeline, sdkComponents, true);
    Job job = translator.translate(pipeline, pipelineProto, sdkComponents, runner, Collections.emptyList()).getJob();
    assertAllStepOutputsHaveUniqueIds(job);
    List<Step> steps = job.getSteps();
    assertEquals(3, steps.size());
    Map<String, Object> parDo1Properties = steps.get(1).getProperties();
    Map<String, Object> parDo2Properties = steps.get(2).getProperties();
    assertThat(parDo1Properties, hasKey("display_data"));
    @SuppressWarnings("unchecked") Collection<Map<String, String>> fn1displayData = (Collection<Map<String, String>>) parDo1Properties.get("display_data");
    @SuppressWarnings("unchecked") Collection<Map<String, String>> fn2displayData = (Collection<Map<String, String>>) parDo2Properties.get("display_data");
    ImmutableSet<ImmutableMap<String, Object>> expectedFn1DisplayData = ImmutableSet.of(ImmutableMap.<String, Object>builder().put("key", "foo").put("type", "STRING").put("value", "bar").put("namespace", fn1.getClass().getName()).build(), ImmutableMap.<String, Object>builder().put("key", "fn").put("label", "Transform Function").put("type", "JAVA_CLASS").put("value", fn1.getClass().getName()).put("shortValue", fn1.getClass().getSimpleName()).put("namespace", parDo1.getClass().getName()).build(), ImmutableMap.<String, Object>builder().put("key", "foo2").put("type", "JAVA_CLASS").put("value", DataflowPipelineTranslatorTest.class.getName()).put("shortValue", DataflowPipelineTranslatorTest.class.getSimpleName()).put("namespace", fn1.getClass().getName()).put("label", "Test Class").put("linkUrl", "http://www.google.com").build());
    ImmutableSet<ImmutableMap<String, Object>> expectedFn2DisplayData = ImmutableSet.of(ImmutableMap.<String, Object>builder().put("key", "fn").put("label", "Transform Function").put("type", "JAVA_CLASS").put("value", fn2.getClass().getName()).put("shortValue", fn2.getClass().getSimpleName()).put("namespace", parDo2.getClass().getName()).build(), ImmutableMap.<String, Object>builder().put("key", "foo3").put("type", "INTEGER").put("value", 1234L).put("namespace", fn2.getClass().getName()).build());
    assertEquals(expectedFn1DisplayData, ImmutableSet.copyOf(fn1displayData));
    assertEquals(expectedFn2DisplayData, ImmutableSet.copyOf(fn2displayData));
}

Example 85 with Pipeline

use of org.apache.beam.model.pipeline.v1.RunnerApi.Pipeline in project beam by apache.

the class DataflowPipelineTranslatorTest method testInaccessibleProvider.

@Test
public void testInaccessibleProvider() throws Exception {
    DataflowPipelineOptions options = buildPipelineOptions();
    Pipeline pipeline = Pipeline.create(options);
    DataflowPipelineTranslator t = DataflowPipelineTranslator.fromOptions(options);
    pipeline.apply(TextIO.read().from(new TestValueProvider()));
    // Check that translation does not fail.
    SdkComponents sdkComponents = createSdkComponents(options);
    RunnerApi.Pipeline pipelineProto = PipelineTranslation.toProto(pipeline, sdkComponents, true);
    t.translate(pipeline, pipelineProto, sdkComponents, DataflowRunner.fromOptions(options), Collections.emptyList());
}