Search in sources :

Example 11 with WindowPane

use of org.apache.samza.operators.windows.WindowPane in project samza by apache.

the class TestWindowOperator method testCancellationOfOnceTrigger.

@Test
public void testCancellationOfOnceTrigger() throws Exception {
    OperatorSpecGraph sgb = this.getKeyedTumblingWindowStreamGraph(AccumulationMode.ACCUMULATING, Duration.ofSeconds(1), Triggers.count(2)).getOperatorSpecGraph();
    TestClock testClock = new TestClock();
    StreamOperatorTask task = new StreamOperatorTask(sgb, testClock);
    task.init(this.context);
    List<WindowPane<Integer, Collection<IntegerEnvelope>>> windowPanes = new ArrayList<>();
    MessageCollector messageCollector = envelope -> windowPanes.add((WindowPane<Integer, Collection<IntegerEnvelope>>) envelope.getMessage());
    task.processAsync(new IntegerEnvelope(1), messageCollector, taskCoordinator, taskCallback);
    task.processAsync(new IntegerEnvelope(1), messageCollector, taskCoordinator, taskCallback);
    Assert.assertEquals(windowPanes.size(), 1);
    Assert.assertEquals(windowPanes.get(0).getKey().getPaneId(), "0");
    Assert.assertEquals(windowPanes.get(0).getKey().getKey(), new Integer(1));
    Assert.assertEquals(windowPanes.get(0).getFiringType(), FiringType.EARLY);
    task.processAsync(new IntegerEnvelope(1), messageCollector, taskCoordinator, taskCallback);
    task.processAsync(new IntegerEnvelope(1), messageCollector, taskCoordinator, taskCallback);
    task.processAsync(new IntegerEnvelope(1), messageCollector, taskCoordinator, taskCallback);
    Assert.assertEquals(windowPanes.size(), 1);
    testClock.advanceTime(Duration.ofSeconds(1));
    task.window(messageCollector, taskCoordinator);
    Assert.assertEquals(windowPanes.size(), 2);
    Assert.assertEquals(windowPanes.get(0).getKey().getPaneId(), "0");
    Assert.assertEquals(windowPanes.get(1).getKey().getPaneId(), "0");
    Assert.assertEquals(windowPanes.get(1).getFiringType(), FiringType.DEFAULT);
    task.processAsync(new IntegerEnvelope(3), messageCollector, taskCoordinator, taskCallback);
    testClock.advanceTime(Duration.ofSeconds(1));
    task.window(messageCollector, taskCoordinator);
    Assert.assertEquals(windowPanes.size(), 3);
    Assert.assertEquals(windowPanes.get(2).getKey().getKey(), new Integer(3));
    Assert.assertEquals(windowPanes.get(2).getKey().getPaneId(), "1000");
    Assert.assertEquals(windowPanes.get(2).getFiringType(), FiringType.DEFAULT);
    Assert.assertEquals((windowPanes.get(2).getMessage()).size(), 1);
}
Also used : TestClock(org.apache.samza.testUtils.TestClock) StreamApplicationDescriptorImpl(org.apache.samza.application.descriptors.StreamApplicationDescriptorImpl) TaskModel(org.apache.samza.job.model.TaskModel) GenericInputDescriptor(org.apache.samza.system.descriptors.GenericInputDescriptor) OperatorSpecGraph(org.apache.samza.operators.OperatorSpecGraph) AccumulationMode(org.apache.samza.operators.windows.AccumulationMode) Duration(java.time.Duration) Map(java.util.Map) MapConfig(org.apache.samza.config.MapConfig) KV(org.apache.samza.operators.KV) TaskCallback(org.apache.samza.task.TaskCallback) ImmutableSet(com.google.common.collect.ImmutableSet) TaskName(org.apache.samza.container.TaskName) IncomingMessageEnvelope(org.apache.samza.system.IncomingMessageEnvelope) Collection(java.util.Collection) Triggers(org.apache.samza.operators.triggers.Triggers) Context(org.apache.samza.context.Context) List(java.util.List) TestClock(org.apache.samza.testUtils.TestClock) TestInMemoryStore(org.apache.samza.operators.impl.store.TestInMemoryStore) Config(org.apache.samza.config.Config) KVSerde(org.apache.samza.serializers.KVSerde) StreamApplication(org.apache.samza.application.StreamApplication) MetricsRegistryMap(org.apache.samza.metrics.MetricsRegistryMap) Mockito.mock(org.mockito.Mockito.mock) StreamOperatorTask(org.apache.samza.task.StreamOperatorTask) GenericSystemDescriptor(org.apache.samza.system.descriptors.GenericSystemDescriptor) HashMap(java.util.HashMap) Serde(org.apache.samza.serializers.Serde) SystemStreamPartition(org.apache.samza.system.SystemStreamPartition) ArrayList(java.util.ArrayList) MapFunction(org.apache.samza.operators.functions.MapFunction) WindowPane(org.apache.samza.operators.windows.WindowPane) Trigger(org.apache.samza.operators.triggers.Trigger) ImmutableList(com.google.common.collect.ImmutableList) MessageCollector(org.apache.samza.task.MessageCollector) FiringType(org.apache.samza.operators.triggers.FiringType) SystemStream(org.apache.samza.system.SystemStream) MockContext(org.apache.samza.context.MockContext) IntegerSerde(org.apache.samza.serializers.IntegerSerde) MessageStream(org.apache.samza.operators.MessageStream) Before(org.junit.Before) Windows(org.apache.samza.operators.windows.Windows) Partition(org.apache.samza.Partition) IOException(java.io.IOException) Test(org.junit.Test) Mockito.times(org.mockito.Mockito.times) Mockito.when(org.mockito.Mockito.when) TaskCoordinator(org.apache.samza.task.TaskCoordinator) Mockito.verify(org.mockito.Mockito.verify) OutgoingMessageEnvelope(org.apache.samza.system.OutgoingMessageEnvelope) TaskContextImpl(org.apache.samza.context.TaskContextImpl) Assert(org.junit.Assert) TimeSeriesKeySerde(org.apache.samza.operators.impl.store.TimeSeriesKeySerde) OperatorSpecGraph(org.apache.samza.operators.OperatorSpecGraph) StreamOperatorTask(org.apache.samza.task.StreamOperatorTask) MessageCollector(org.apache.samza.task.MessageCollector) ArrayList(java.util.ArrayList) Collection(java.util.Collection) WindowPane(org.apache.samza.operators.windows.WindowPane) Test(org.junit.Test)

Example 12 with WindowPane

use of org.apache.samza.operators.windows.WindowPane in project samza by apache.

the class WindowExample method describe.

@Override
public void describe(StreamApplicationDescriptor appDescriptor) {
    KafkaSystemDescriptor trackingSystem = new KafkaSystemDescriptor("tracking");
    KafkaInputDescriptor<PageViewEvent> inputStreamDescriptor = trackingSystem.getInputDescriptor("pageViewEvent", new JsonSerdeV2<>(PageViewEvent.class));
    KafkaOutputDescriptor<Integer> outputStreamDescriptor = trackingSystem.getOutputDescriptor("pageViewEventPerMember", new IntegerSerde());
    SupplierFunction<Integer> initialValue = () -> 0;
    FoldLeftFunction<PageViewEvent, Integer> counter = (m, c) -> c == null ? 1 : c + 1;
    MessageStream<PageViewEvent> inputStream = appDescriptor.getInputStream(inputStreamDescriptor);
    OutputStream<Integer> outputStream = appDescriptor.getOutputStream(outputStreamDescriptor);
    // create a tumbling window that outputs the number of message collected every 10 minutes.
    // also emit early results if either the number of messages collected reaches 30000, or if no new messages arrive
    // for 1 minute.
    inputStream.window(Windows.tumblingWindow(Duration.ofMinutes(10), initialValue, counter, new IntegerSerde()).setLateTrigger(Triggers.any(Triggers.count(30000), Triggers.timeSinceLastMessage(Duration.ofMinutes(1)))), "window").map(WindowPane::getMessage).sendTo(outputStream);
}
Also used : ApplicationRunner(org.apache.samza.runtime.ApplicationRunner) Windows(org.apache.samza.operators.windows.Windows) KafkaInputDescriptor(org.apache.samza.system.kafka.descriptors.KafkaInputDescriptor) CommandLine(org.apache.samza.util.CommandLine) KafkaSystemDescriptor(org.apache.samza.system.kafka.descriptors.KafkaSystemDescriptor) PageViewEvent(org.apache.samza.example.models.PageViewEvent) Triggers(org.apache.samza.operators.triggers.Triggers) WindowPane(org.apache.samza.operators.windows.WindowPane) KafkaOutputDescriptor(org.apache.samza.system.kafka.descriptors.KafkaOutputDescriptor) StreamApplicationDescriptor(org.apache.samza.application.descriptors.StreamApplicationDescriptor) Duration(java.time.Duration) Config(org.apache.samza.config.Config) ApplicationRunners(org.apache.samza.runtime.ApplicationRunners) JsonSerdeV2(org.apache.samza.serializers.JsonSerdeV2) StreamApplication(org.apache.samza.application.StreamApplication) OutputStream(org.apache.samza.operators.OutputStream) FoldLeftFunction(org.apache.samza.operators.functions.FoldLeftFunction) IntegerSerde(org.apache.samza.serializers.IntegerSerde) SupplierFunction(org.apache.samza.operators.functions.SupplierFunction) MessageStream(org.apache.samza.operators.MessageStream) PageViewEvent(org.apache.samza.example.models.PageViewEvent) KafkaSystemDescriptor(org.apache.samza.system.kafka.descriptors.KafkaSystemDescriptor) IntegerSerde(org.apache.samza.serializers.IntegerSerde)

Example 13 with WindowPane

use of org.apache.samza.operators.windows.WindowPane in project samza by apache.

the class PageViewCounterExample method describe.

@Override
public void describe(StreamApplicationDescriptor appDescriptor) {
    KafkaSystemDescriptor trackingSystem = new KafkaSystemDescriptor("tracking");
    KafkaInputDescriptor<PageViewEvent> inputStreamDescriptor = trackingSystem.getInputDescriptor("pageViewEvent", new JsonSerdeV2<>(PageViewEvent.class));
    KafkaOutputDescriptor<KV<String, PageViewCount>> outputStreamDescriptor = trackingSystem.getOutputDescriptor("pageViewEventPerMember", KVSerde.of(new StringSerde(), new JsonSerdeV2<>(PageViewCount.class)));
    MessageStream<PageViewEvent> pageViewEvents = appDescriptor.getInputStream(inputStreamDescriptor);
    OutputStream<KV<String, PageViewCount>> pageViewEventPerMemberStream = appDescriptor.getOutputStream(outputStreamDescriptor);
    SupplierFunction<Integer> initialValue = () -> 0;
    FoldLeftFunction<PageViewEvent, Integer> foldLeftFn = (m, c) -> c + 1;
    pageViewEvents.window(Windows.keyedTumblingWindow(PageViewEvent::getMemberId, Duration.ofSeconds(10), initialValue, foldLeftFn, null, null).setEarlyTrigger(Triggers.repeat(Triggers.count(5))).setAccumulationMode(AccumulationMode.DISCARDING), "tumblingWindow").map(windowPane -> KV.of(windowPane.getKey().getKey(), buildPageViewCount(windowPane))).sendTo(pageViewEventPerMemberStream);
}
Also used : ApplicationRunner(org.apache.samza.runtime.ApplicationRunner) Windows(org.apache.samza.operators.windows.Windows) KafkaInputDescriptor(org.apache.samza.system.kafka.descriptors.KafkaInputDescriptor) CommandLine(org.apache.samza.util.CommandLine) KafkaSystemDescriptor(org.apache.samza.system.kafka.descriptors.KafkaSystemDescriptor) PageViewEvent(org.apache.samza.example.models.PageViewEvent) Triggers(org.apache.samza.operators.triggers.Triggers) WindowPane(org.apache.samza.operators.windows.WindowPane) StringSerde(org.apache.samza.serializers.StringSerde) PageViewCount(org.apache.samza.example.models.PageViewCount) KafkaOutputDescriptor(org.apache.samza.system.kafka.descriptors.KafkaOutputDescriptor) StreamApplicationDescriptor(org.apache.samza.application.descriptors.StreamApplicationDescriptor) AccumulationMode(org.apache.samza.operators.windows.AccumulationMode) Duration(java.time.Duration) Config(org.apache.samza.config.Config) ApplicationRunners(org.apache.samza.runtime.ApplicationRunners) JsonSerdeV2(org.apache.samza.serializers.JsonSerdeV2) KVSerde(org.apache.samza.serializers.KVSerde) StreamApplication(org.apache.samza.application.StreamApplication) KV(org.apache.samza.operators.KV) OutputStream(org.apache.samza.operators.OutputStream) FoldLeftFunction(org.apache.samza.operators.functions.FoldLeftFunction) SupplierFunction(org.apache.samza.operators.functions.SupplierFunction) MessageStream(org.apache.samza.operators.MessageStream) StringSerde(org.apache.samza.serializers.StringSerde) PageViewEvent(org.apache.samza.example.models.PageViewEvent) KV(org.apache.samza.operators.KV) KafkaSystemDescriptor(org.apache.samza.system.kafka.descriptors.KafkaSystemDescriptor) JsonSerdeV2(org.apache.samza.serializers.JsonSerdeV2)

Example 14 with WindowPane

use of org.apache.samza.operators.windows.WindowPane in project samza by apache.

the class TestWindowOperator method testTumblingWindowsDiscardingMode.

@Test
public void testTumblingWindowsDiscardingMode() throws Exception {
    StreamApplication sgb = new KeyedTumblingWindowStreamApplication(AccumulationMode.DISCARDING, Duration.ofSeconds(1), Triggers.repeat(Triggers.count(2)));
    List<WindowPane<Integer, Collection<IntegerEnvelope>>> windowPanes = new ArrayList<>();
    TestClock testClock = new TestClock();
    StreamOperatorTask task = new StreamOperatorTask(sgb, runner, testClock);
    task.init(config, taskContext);
    MessageCollector messageCollector = envelope -> windowPanes.add((WindowPane<Integer, Collection<IntegerEnvelope>>) envelope.getMessage());
    integers.forEach(n -> task.process(new IntegerEnvelope(n), messageCollector, taskCoordinator));
    testClock.advanceTime(Duration.ofSeconds(1));
    task.window(messageCollector, taskCoordinator);
    Assert.assertEquals(windowPanes.size(), 5);
    Assert.assertEquals(windowPanes.get(0).getKey().getKey(), new Integer(1));
    Assert.assertEquals((windowPanes.get(0).getMessage()).size(), 2);
    Assert.assertEquals(windowPanes.get(1).getKey().getKey(), new Integer(2));
    Assert.assertEquals((windowPanes.get(1).getMessage()).size(), 2);
    Assert.assertEquals(windowPanes.get(2).getKey().getKey(), new Integer(1));
    Assert.assertEquals((windowPanes.get(2).getMessage()).size(), 2);
    Assert.assertEquals(windowPanes.get(3).getKey().getKey(), new Integer(2));
    Assert.assertEquals((windowPanes.get(3).getMessage()).size(), 2);
    Assert.assertEquals(windowPanes.get(4).getKey().getKey(), new Integer(3));
    Assert.assertEquals((windowPanes.get(4).getMessage()).size(), 1);
}
Also used : TestClock(org.apache.samza.testUtils.TestClock) StreamOperatorTask(org.apache.samza.task.StreamOperatorTask) SystemStreamPartition(org.apache.samza.system.SystemStreamPartition) Function(java.util.function.Function) WindowPane(org.apache.samza.operators.windows.WindowPane) ArrayList(java.util.ArrayList) Trigger(org.apache.samza.operators.triggers.Trigger) ImmutableList(com.google.common.collect.ImmutableList) AccumulationMode(org.apache.samza.operators.windows.AccumulationMode) MessageCollector(org.apache.samza.task.MessageCollector) FiringType(org.apache.samza.operators.triggers.FiringType) SystemStream(org.apache.samza.system.SystemStream) Duration(java.time.Duration) Before(org.junit.Before) ApplicationRunner(org.apache.samza.runtime.ApplicationRunner) TaskContext(org.apache.samza.task.TaskContext) ImmutableSet(com.google.common.collect.ImmutableSet) Windows(org.apache.samza.operators.windows.Windows) IncomingMessageEnvelope(org.apache.samza.system.IncomingMessageEnvelope) Collection(java.util.Collection) Partition(org.apache.samza.Partition) StreamSpec(org.apache.samza.system.StreamSpec) Test(org.junit.Test) Mockito.when(org.mockito.Mockito.when) TaskCoordinator(org.apache.samza.task.TaskCoordinator) Triggers(org.apache.samza.operators.triggers.Triggers) List(java.util.List) TestClock(org.apache.samza.testUtils.TestClock) Assert(junit.framework.Assert) OutgoingMessageEnvelope(org.apache.samza.system.OutgoingMessageEnvelope) Config(org.apache.samza.config.Config) StreamApplication(org.apache.samza.application.StreamApplication) MetricsRegistryMap(org.apache.samza.metrics.MetricsRegistryMap) Mockito.mock(org.mockito.Mockito.mock) StreamApplication(org.apache.samza.application.StreamApplication) StreamOperatorTask(org.apache.samza.task.StreamOperatorTask) MessageCollector(org.apache.samza.task.MessageCollector) ArrayList(java.util.ArrayList) Collection(java.util.Collection) WindowPane(org.apache.samza.operators.windows.WindowPane) Test(org.junit.Test)

Example 15 with WindowPane

use of org.apache.samza.operators.windows.WindowPane in project samza by apache.

the class TestWindowOperator method testSessionWindowsDiscardingMode.

@Test
public void testSessionWindowsDiscardingMode() throws Exception {
    StreamApplication sgb = new KeyedSessionWindowStreamApplication(AccumulationMode.DISCARDING, Duration.ofMillis(500));
    TestClock testClock = new TestClock();
    List<WindowPane<Integer, Collection<IntegerEnvelope>>> windowPanes = new ArrayList<>();
    StreamOperatorTask task = new StreamOperatorTask(sgb, runner, testClock);
    task.init(config, taskContext);
    MessageCollector messageCollector = envelope -> windowPanes.add((WindowPane<Integer, Collection<IntegerEnvelope>>) envelope.getMessage());
    task.process(new IntegerEnvelope(1), messageCollector, taskCoordinator);
    task.process(new IntegerEnvelope(1), messageCollector, taskCoordinator);
    testClock.advanceTime(Duration.ofSeconds(1));
    task.window(messageCollector, taskCoordinator);
    Assert.assertEquals(windowPanes.size(), 1);
    Assert.assertEquals(windowPanes.get(0).getKey().getPaneId(), "1");
    Assert.assertEquals(windowPanes.get(0).getKey().getKey(), new Integer(1));
    task.process(new IntegerEnvelope(2), messageCollector, taskCoordinator);
    task.process(new IntegerEnvelope(2), messageCollector, taskCoordinator);
    task.process(new IntegerEnvelope(3), messageCollector, taskCoordinator);
    task.process(new IntegerEnvelope(3), messageCollector, taskCoordinator);
    testClock.advanceTime(Duration.ofSeconds(1));
    task.window(messageCollector, taskCoordinator);
    Assert.assertEquals(windowPanes.size(), 3);
    Assert.assertEquals(windowPanes.get(0).getKey().getPaneId(), "1");
    Assert.assertEquals(windowPanes.get(1).getKey().getPaneId(), "1001");
    Assert.assertEquals(windowPanes.get(2).getKey().getPaneId(), "1001");
    Assert.assertEquals((windowPanes.get(0).getMessage()).size(), 2);
    Assert.assertEquals((windowPanes.get(1).getMessage()).size(), 2);
    Assert.assertEquals((windowPanes.get(2).getMessage()).size(), 2);
    task.process(new IntegerEnvelope(2), messageCollector, taskCoordinator);
    task.process(new IntegerEnvelope(2), messageCollector, taskCoordinator);
    testClock.advanceTime(Duration.ofSeconds(1));
    task.window(messageCollector, taskCoordinator);
    Assert.assertEquals(windowPanes.size(), 4);
    Assert.assertEquals(windowPanes.get(3).getKey().getKey(), new Integer(2));
    Assert.assertEquals(windowPanes.get(3).getKey().getPaneId(), "2001");
    Assert.assertEquals((windowPanes.get(3).getMessage()).size(), 2);
}
Also used : TestClock(org.apache.samza.testUtils.TestClock) StreamOperatorTask(org.apache.samza.task.StreamOperatorTask) SystemStreamPartition(org.apache.samza.system.SystemStreamPartition) Function(java.util.function.Function) WindowPane(org.apache.samza.operators.windows.WindowPane) ArrayList(java.util.ArrayList) Trigger(org.apache.samza.operators.triggers.Trigger) ImmutableList(com.google.common.collect.ImmutableList) AccumulationMode(org.apache.samza.operators.windows.AccumulationMode) MessageCollector(org.apache.samza.task.MessageCollector) FiringType(org.apache.samza.operators.triggers.FiringType) SystemStream(org.apache.samza.system.SystemStream) Duration(java.time.Duration) Before(org.junit.Before) ApplicationRunner(org.apache.samza.runtime.ApplicationRunner) TaskContext(org.apache.samza.task.TaskContext) ImmutableSet(com.google.common.collect.ImmutableSet) Windows(org.apache.samza.operators.windows.Windows) IncomingMessageEnvelope(org.apache.samza.system.IncomingMessageEnvelope) Collection(java.util.Collection) Partition(org.apache.samza.Partition) StreamSpec(org.apache.samza.system.StreamSpec) Test(org.junit.Test) Mockito.when(org.mockito.Mockito.when) TaskCoordinator(org.apache.samza.task.TaskCoordinator) Triggers(org.apache.samza.operators.triggers.Triggers) List(java.util.List) TestClock(org.apache.samza.testUtils.TestClock) Assert(junit.framework.Assert) OutgoingMessageEnvelope(org.apache.samza.system.OutgoingMessageEnvelope) Config(org.apache.samza.config.Config) StreamApplication(org.apache.samza.application.StreamApplication) MetricsRegistryMap(org.apache.samza.metrics.MetricsRegistryMap) Mockito.mock(org.mockito.Mockito.mock) StreamApplication(org.apache.samza.application.StreamApplication) StreamOperatorTask(org.apache.samza.task.StreamOperatorTask) MessageCollector(org.apache.samza.task.MessageCollector) ArrayList(java.util.ArrayList) Collection(java.util.Collection) WindowPane(org.apache.samza.operators.windows.WindowPane) Test(org.junit.Test)

Aggregations

WindowPane (org.apache.samza.operators.windows.WindowPane)26 Duration (java.time.Duration)24 Windows (org.apache.samza.operators.windows.Windows)24 Collection (java.util.Collection)23 StreamApplication (org.apache.samza.application.StreamApplication)23 Config (org.apache.samza.config.Config)23 Test (org.junit.Test)23 Mockito.mock (org.mockito.Mockito.mock)23 Mockito.when (org.mockito.Mockito.when)23 ArrayList (java.util.ArrayList)22 List (java.util.List)22 Triggers (org.apache.samza.operators.triggers.Triggers)22 OutgoingMessageEnvelope (org.apache.samza.system.OutgoingMessageEnvelope)22 SystemStream (org.apache.samza.system.SystemStream)22 MessageCollector (org.apache.samza.task.MessageCollector)22 TaskCoordinator (org.apache.samza.task.TaskCoordinator)22 ImmutableList (com.google.common.collect.ImmutableList)21 AccumulationMode (org.apache.samza.operators.windows.AccumulationMode)21 ImmutableSet (com.google.common.collect.ImmutableSet)20 Partition (org.apache.samza.Partition)20