Search in sources :

Example 16 with Node

use of org.apache.beam.runners.dataflow.worker.graph.Nodes.Node in project beam by apache.

the class DeduceNodeLocationsFunctionTest method testSingleNodeWithSdkRead.

@Test
public void testSingleNodeWithSdkRead() throws Exception {
    Node unknown = createReadNode("Unknown", CUSTOM_SOURCE);
    MutableNetwork<Node, Edge> network = createEmptyNetwork();
    network.addNode(unknown);
    Network<Node, Edge> inputNetwork = ImmutableNetwork.copyOf(network);
    network = new DeduceNodeLocationsFunction().apply(network);
    assertThatNetworksAreIdentical(inputNetwork, network);
    for (Node node : ImmutableList.copyOf(network.nodes())) {
        assertNodesIdenticalExceptForExecutionLocation(unknown, node);
        assertThatLocationIsProperlyDeduced(node, ExecutionLocation.SDK_HARNESS);
    }
}
Also used : InstructionOutputNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.InstructionOutputNode) ParallelInstructionNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.ParallelInstructionNode) Node(org.apache.beam.runners.dataflow.worker.graph.Nodes.Node) Edge(org.apache.beam.runners.dataflow.worker.graph.Edges.Edge) DefaultEdge(org.apache.beam.runners.dataflow.worker.graph.Edges.DefaultEdge) Test(org.junit.Test)

Example 17 with Node

use of org.apache.beam.runners.dataflow.worker.graph.Nodes.Node in project beam by apache.

the class InsertFetchAndFilterStreamingSideInputNodesTest method testSdkParDoWithSideInput.

@Test
public void testSdkParDoWithSideInput() throws Exception {
    Pipeline p = Pipeline.create();
    PCollection<String> pc = p.apply(Create.of("a", "b", "c"));
    PCollectionView<List<String>> pcView = pc.apply(View.asList());
    pc.apply(ParDo.of(new TestDoFn()).withSideInputs(pcView));
    RunnerApi.Pipeline pipeline = PipelineTranslation.toProto(p);
    Node predecessor = createParDoNode("predecessor");
    InstructionOutputNode mainInput = InstructionOutputNode.create(new InstructionOutput(), "fakeId");
    Node sideInputParDo = createParDoNode(findParDoWithSideInput(pipeline));
    MutableNetwork<Node, Edge> network = createEmptyNetwork();
    network.addNode(predecessor);
    network.addNode(mainInput);
    network.addNode(sideInputParDo);
    network.addEdge(predecessor, mainInput, DefaultEdge.create());
    network.addEdge(mainInput, sideInputParDo, DefaultEdge.create());
    network = InsertFetchAndFilterStreamingSideInputNodes.with(pipeline).forNetwork(network);
    Node mainInputClone = InstructionOutputNode.create(mainInput.getInstructionOutput(), "fakeId");
    Node fetchAndFilter = FetchAndFilterStreamingSideInputsNode.create(pcView.getWindowingStrategyInternal(), ImmutableMap.of(pcView, ParDoTranslation.translateWindowMappingFn(pcView.getWindowMappingFn(), SdkComponents.create(PipelineOptionsFactory.create()))), NameContextsForTests.nameContextForTest());
    MutableNetwork<Node, Edge> expectedNetwork = createEmptyNetwork();
    expectedNetwork.addNode(predecessor);
    expectedNetwork.addNode(mainInputClone);
    expectedNetwork.addNode(fetchAndFilter);
    expectedNetwork.addNode(mainInput);
    expectedNetwork.addNode(sideInputParDo);
    expectedNetwork.addEdge(predecessor, mainInputClone, DefaultEdge.create());
    expectedNetwork.addEdge(mainInputClone, fetchAndFilter, DefaultEdge.create());
    expectedNetwork.addEdge(fetchAndFilter, mainInput, DefaultEdge.create());
    expectedNetwork.addEdge(mainInput, sideInputParDo, DefaultEdge.create());
    assertThatNetworksAreIdentical(expectedNetwork, network);
}
Also used : FetchAndFilterStreamingSideInputsNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.FetchAndFilterStreamingSideInputsNode) InstructionOutputNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.InstructionOutputNode) ParallelInstructionNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.ParallelInstructionNode) Node(org.apache.beam.runners.dataflow.worker.graph.Nodes.Node) InstructionOutput(com.google.api.services.dataflow.model.InstructionOutput) Pipeline(org.apache.beam.sdk.Pipeline) RunnerApi(org.apache.beam.model.pipeline.v1.RunnerApi) InstructionOutputNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.InstructionOutputNode) ArrayList(java.util.ArrayList) List(java.util.List) Edge(org.apache.beam.runners.dataflow.worker.graph.Edges.Edge) DefaultEdge(org.apache.beam.runners.dataflow.worker.graph.Edges.DefaultEdge) Test(org.junit.Test)

Example 18 with Node

use of org.apache.beam.runners.dataflow.worker.graph.Nodes.Node in project beam by apache.

the class InsertFetchAndFilterStreamingSideInputNodesTest method allPathsWithWrappedNodes.

private List<List<Equivalence.Wrapper<Node>>> allPathsWithWrappedNodes(Network<Node, Edge> network) {
    List<List<Node>> paths = Networks.allPathsFromRootsToLeaves(network);
    List<List<Equivalence.Wrapper<Node>>> wrappedPaths = new ArrayList<>();
    for (List<Node> path : paths) {
        List<Equivalence.Wrapper<Node>> wrappedPath = new ArrayList<>();
        for (Node node : path) {
            wrappedPath.add(NodeEquivalence.INSTANCE.wrap(node));
        }
        wrappedPaths.add(wrappedPath);
    }
    return wrappedPaths;
}
Also used : Wrapper(org.apache.beam.vendor.guava.v26_0_jre.com.google.common.base.Equivalence.Wrapper) Equivalence(org.apache.beam.vendor.guava.v26_0_jre.com.google.common.base.Equivalence) FetchAndFilterStreamingSideInputsNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.FetchAndFilterStreamingSideInputsNode) InstructionOutputNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.InstructionOutputNode) ParallelInstructionNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.ParallelInstructionNode) Node(org.apache.beam.runners.dataflow.worker.graph.Nodes.Node) ArrayList(java.util.ArrayList) ArrayList(java.util.ArrayList) List(java.util.List)

Example 19 with Node

use of org.apache.beam.runners.dataflow.worker.graph.Nodes.Node in project beam by apache.

the class InsertFetchAndFilterStreamingSideInputNodesTest method testWithoutPipeline.

@Test
public void testWithoutPipeline() throws Exception {
    Node unknown = createParDoNode("parDoId");
    MutableNetwork<Node, Edge> network = createEmptyNetwork();
    network.addNode(unknown);
    Network<Node, Edge> inputNetwork = ImmutableNetwork.copyOf(network);
    network = InsertFetchAndFilterStreamingSideInputNodes.with(null).forNetwork(network);
    assertThatNetworksAreIdentical(inputNetwork, network);
}
Also used : FetchAndFilterStreamingSideInputsNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.FetchAndFilterStreamingSideInputsNode) InstructionOutputNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.InstructionOutputNode) ParallelInstructionNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.ParallelInstructionNode) Node(org.apache.beam.runners.dataflow.worker.graph.Nodes.Node) Edge(org.apache.beam.runners.dataflow.worker.graph.Edges.Edge) DefaultEdge(org.apache.beam.runners.dataflow.worker.graph.Edges.DefaultEdge) Test(org.junit.Test)

Example 20 with Node

use of org.apache.beam.runners.dataflow.worker.graph.Nodes.Node in project beam by apache.

the class MapTaskToNetworkFunctionTest method testMultipleOutput.

@Test
public void testMultipleOutput() {
    // /---> WriteA
    // Read ---> ParDo
    // \---> WriteB
    InstructionOutput readOutput = createInstructionOutput("Read.out");
    ParallelInstruction read = createParallelInstruction("Read", readOutput);
    read.setRead(new ReadInstruction());
    MultiOutputInfo parDoMultiOutput1 = createMultiOutputInfo("output1");
    MultiOutputInfo parDoMultiOutput2 = createMultiOutputInfo("output2");
    ParDoInstruction parDoInstruction = new ParDoInstruction();
    // Read.out
    parDoInstruction.setInput(createInstructionInput(0, 0));
    parDoInstruction.setMultiOutputInfos(ImmutableList.of(parDoMultiOutput1, parDoMultiOutput2));
    InstructionOutput parDoOutput1 = createInstructionOutput("ParDo.out1");
    InstructionOutput parDoOutput2 = createInstructionOutput("ParDo.out2");
    ParallelInstruction parDo = createParallelInstruction("ParDo", parDoOutput1, parDoOutput2);
    parDo.setParDo(parDoInstruction);
    WriteInstruction writeAInstruction = new WriteInstruction();
    // ParDo.out1
    writeAInstruction.setInput(createInstructionInput(1, 0));
    ParallelInstruction writeA = createParallelInstruction("WriteA");
    writeA.setWrite(writeAInstruction);
    WriteInstruction writeBInstruction = new WriteInstruction();
    // ParDo.out2
    writeBInstruction.setInput(createInstructionInput(1, 1));
    ParallelInstruction writeB = createParallelInstruction("WriteB");
    writeB.setWrite(writeBInstruction);
    MapTask mapTask = new MapTask();
    mapTask.setInstructions(ImmutableList.of(read, parDo, writeA, writeB));
    mapTask.setFactory(Transport.getJsonFactory());
    Network<Node, Edge> network = new MapTaskToNetworkFunction(IdGenerators.decrementingLongs()).apply(mapTask);
    assertNetworkProperties(network);
    assertEquals(7, network.nodes().size());
    assertEquals(6, network.edges().size());
    ParallelInstructionNode parDoNode = get(network, parDo);
    ParallelInstructionNode writeANode = get(network, writeA);
    ParallelInstructionNode writeBNode = get(network, writeB);
    InstructionOutputNode parDoOutput1Node = getOnlyPredecessor(network, writeANode);
    assertEquals(parDoOutput1, parDoOutput1Node.getInstructionOutput());
    InstructionOutputNode parDoOutput2Node = getOnlyPredecessor(network, writeBNode);
    assertEquals(parDoOutput2, parDoOutput2Node.getInstructionOutput());
    assertThat(network.successors(parDoNode), Matchers.<Node>containsInAnyOrder(parDoOutput1Node, parDoOutput2Node));
    assertEquals(parDoMultiOutput1, ((MultiOutputInfoEdge) Iterables.getOnlyElement(network.edgesConnecting(parDoNode, parDoOutput1Node))).getMultiOutputInfo());
    assertEquals(parDoMultiOutput2, ((MultiOutputInfoEdge) Iterables.getOnlyElement(network.edgesConnecting(parDoNode, parDoOutput2Node))).getMultiOutputInfo());
}
Also used : MultiOutputInfo(com.google.api.services.dataflow.model.MultiOutputInfo) InstructionOutputNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.InstructionOutputNode) ParallelInstructionNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.ParallelInstructionNode) Node(org.apache.beam.runners.dataflow.worker.graph.Nodes.Node) InstructionOutput(com.google.api.services.dataflow.model.InstructionOutput) ParallelInstructionNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.ParallelInstructionNode) ReadInstruction(com.google.api.services.dataflow.model.ReadInstruction) ParallelInstruction(com.google.api.services.dataflow.model.ParallelInstruction) ParDoInstruction(com.google.api.services.dataflow.model.ParDoInstruction) InstructionOutputNode(org.apache.beam.runners.dataflow.worker.graph.Nodes.InstructionOutputNode) MapTask(com.google.api.services.dataflow.model.MapTask) WriteInstruction(com.google.api.services.dataflow.model.WriteInstruction) Edge(org.apache.beam.runners.dataflow.worker.graph.Edges.Edge) DefaultEdge(org.apache.beam.runners.dataflow.worker.graph.Edges.DefaultEdge) MultiOutputInfoEdge(org.apache.beam.runners.dataflow.worker.graph.Edges.MultiOutputInfoEdge) Test(org.junit.Test)

Aggregations

Node (org.apache.beam.runners.dataflow.worker.graph.Nodes.Node)65 ParallelInstructionNode (org.apache.beam.runners.dataflow.worker.graph.Nodes.ParallelInstructionNode)64 InstructionOutputNode (org.apache.beam.runners.dataflow.worker.graph.Nodes.InstructionOutputNode)59 Edge (org.apache.beam.runners.dataflow.worker.graph.Edges.Edge)50 DefaultEdge (org.apache.beam.runners.dataflow.worker.graph.Edges.DefaultEdge)41 Test (org.junit.Test)40 ParallelInstruction (com.google.api.services.dataflow.model.ParallelInstruction)22 MultiOutputInfoEdge (org.apache.beam.runners.dataflow.worker.graph.Edges.MultiOutputInfoEdge)21 InstructionOutput (com.google.api.services.dataflow.model.InstructionOutput)17 List (java.util.List)10 OperationNode (org.apache.beam.runners.dataflow.worker.graph.Nodes.OperationNode)10 MapTask (com.google.api.services.dataflow.model.MapTask)9 ParDoInstruction (com.google.api.services.dataflow.model.ParDoInstruction)9 ReadInstruction (com.google.api.services.dataflow.model.ReadInstruction)9 RemoteGrpcPortNode (org.apache.beam.runners.dataflow.worker.graph.Nodes.RemoteGrpcPortNode)8 FlattenInstruction (com.google.api.services.dataflow.model.FlattenInstruction)7 MultiOutputInfo (com.google.api.services.dataflow.model.MultiOutputInfo)7 ArrayList (java.util.ArrayList)7 HappensBeforeEdge (org.apache.beam.runners.dataflow.worker.graph.Edges.HappensBeforeEdge)7 ParDoOperation (org.apache.beam.runners.dataflow.worker.util.common.worker.ParDoOperation)7