Search in sources :

Example 46 with TopologyTestDriver

use of org.apache.kafka.streams.TopologyTestDriver in project kafka by apache.

the class KTableKTableForeignKeyJoinIntegrationTest method joinShouldProduceNullsWhenValueHasNonMatchingForeignKey.

@Test
public void joinShouldProduceNullsWhenValueHasNonMatchingForeignKey() {
    final Topology topology = getTopology(streamsConfig, materialized ? "store" : null, leftJoin, rejoin);
    try (final TopologyTestDriver driver = new TopologyTestDriver(topology, streamsConfig)) {
        final TestInputTopic<String, String> right = driver.createInputTopic(RIGHT_TABLE, new StringSerializer(), new StringSerializer());
        final TestInputTopic<String, String> left = driver.createInputTopic(LEFT_TABLE, new StringSerializer(), new StringSerializer());
        final TestOutputTopic<String, String> outputTopic = driver.createOutputTopic(OUTPUT, new StringDeserializer(), new StringDeserializer());
        final KeyValueStore<String, String> store = driver.getKeyValueStore("store");
        left.pipeInput("lhs1", "lhsValue1|rhs1");
        // no output for a new inner join on a non-existent FK
        // the left join of course emits the half-joined output
        assertThat(outputTopic.readKeyValuesToMap(), is(leftJoin ? mkMap(mkEntry("lhs1", "(lhsValue1|rhs1,null)")) : emptyMap()));
        if (materialized) {
            assertThat(asMap(store), is(leftJoin ? mkMap(mkEntry("lhs1", "(lhsValue1|rhs1,null)")) : emptyMap()));
        }
        // "moving" our subscription to another non-existent FK results in an unnecessary tombstone for inner join,
        // since it impossible to know whether the prior FK existed or not (and thus whether any results have
        // previously been emitted)
        // The left join emits a _necessary_ update (since the lhs record has actually changed)
        left.pipeInput("lhs1", "lhsValue1|rhs2");
        assertThat(outputTopic.readKeyValuesToMap(), is(mkMap(mkEntry("lhs1", leftJoin ? "(lhsValue1|rhs2,null)" : null))));
        if (materialized) {
            assertThat(asMap(store), is(leftJoin ? mkMap(mkEntry("lhs1", "(lhsValue1|rhs2,null)")) : emptyMap()));
        }
        // of course, moving it again to yet another non-existent FK has the same effect
        left.pipeInput("lhs1", "lhsValue1|rhs3");
        assertThat(outputTopic.readKeyValuesToMap(), is(mkMap(mkEntry("lhs1", leftJoin ? "(lhsValue1|rhs3,null)" : null))));
        if (materialized) {
            assertThat(asMap(store), is(leftJoin ? mkMap(mkEntry("lhs1", "(lhsValue1|rhs3,null)")) : emptyMap()));
        }
        // Adding an RHS record now, so that we can demonstrate "moving" from a non-existent FK to an existent one
        // This RHS key was previously referenced, but it's not referenced now, so adding this record should
        // result in no changes whatsoever.
        right.pipeInput("rhs1", "rhsValue1");
        assertThat(outputTopic.readKeyValuesToMap(), is(emptyMap()));
        if (materialized) {
            assertThat(asMap(store), is(leftJoin ? mkMap(mkEntry("lhs1", "(lhsValue1|rhs3,null)")) : emptyMap()));
        }
        // now, we change to a FK that exists, and see the join completes
        left.pipeInput("lhs1", "lhsValue1|rhs1");
        assertThat(outputTopic.readKeyValuesToMap(), is(mkMap(mkEntry("lhs1", "(lhsValue1|rhs1,rhsValue1)"))));
        if (materialized) {
            assertThat(asMap(store), is(mkMap(mkEntry("lhs1", "(lhsValue1|rhs1,rhsValue1)"))));
        }
        // but if we update it again to a non-existent one, we'll get a tombstone for the inner join, and the
        // left join updates appropriately.
        left.pipeInput("lhs1", "lhsValue1|rhs2");
        assertThat(outputTopic.readKeyValuesToMap(), is(mkMap(mkEntry("lhs1", leftJoin ? "(lhsValue1|rhs2,null)" : null))));
        if (materialized) {
            assertThat(asMap(store), is(leftJoin ? mkMap(mkEntry("lhs1", "(lhsValue1|rhs2,null)")) : emptyMap()));
        }
    }
}
Also used : StringDeserializer(org.apache.kafka.common.serialization.StringDeserializer) TopologyTestDriver(org.apache.kafka.streams.TopologyTestDriver) Topology(org.apache.kafka.streams.Topology) StringSerializer(org.apache.kafka.common.serialization.StringSerializer) Test(org.junit.Test)

Example 47 with TopologyTestDriver

use of org.apache.kafka.streams.TopologyTestDriver in project kafka by apache.

the class ProcessorTopologyTest method testPrefixScanInMemoryStoreWithCachingNoLogging.

@Test
public void testPrefixScanInMemoryStoreWithCachingNoLogging() {
    final StoreBuilder<KeyValueStore<String, String>> storeBuilder = Stores.keyValueStoreBuilder(Stores.inMemoryKeyValueStore(DEFAULT_STORE_NAME), Serdes.String(), Serdes.String()).withCachingEnabled().withLoggingDisabled();
    topology.addSource("source1", STRING_DESERIALIZER, STRING_DESERIALIZER, INPUT_TOPIC_1).addProcessor("processor1", defineWithStores(() -> new StatefulProcessor(DEFAULT_STORE_NAME), Collections.singleton(storeBuilder)), "source1").addSink("counts", OUTPUT_TOPIC_1, "processor1");
    driver = new TopologyTestDriver(topology, props);
    final TestInputTopic<String, String> inputTopic = driver.createInputTopic(INPUT_TOPIC_1, STRING_SERIALIZER, STRING_SERIALIZER);
    final TestOutputTopic<Integer, String> outputTopic1 = driver.createOutputTopic(OUTPUT_TOPIC_1, Serdes.Integer().deserializer(), Serdes.String().deserializer());
    inputTopic.pipeInput("key1", "value1");
    inputTopic.pipeInput("key2", "value2");
    inputTopic.pipeInput("key3", "value3");
    inputTopic.pipeInput("key1", "value4");
    assertTrue(outputTopic1.isEmpty());
    final KeyValueStore<String, String> store = driver.getKeyValueStore(DEFAULT_STORE_NAME);
    final List<KeyValue<String, String>> results = prefixScanResults(store, DEFAULT_PREFIX);
    assertEquals("key1", results.get(0).key);
    assertEquals("value4", results.get(0).value);
    assertEquals("key2", results.get(1).key);
    assertEquals("value2", results.get(1).value);
    assertEquals("key3", results.get(2).key);
    assertEquals("value3", results.get(2).value);
}
Also used : KeyValue(org.apache.kafka.streams.KeyValue) TopologyTestDriver(org.apache.kafka.streams.TopologyTestDriver) KeyValueStore(org.apache.kafka.streams.state.KeyValueStore) CoreMatchers.containsString(org.hamcrest.CoreMatchers.containsString) Test(org.junit.Test)

Example 48 with TopologyTestDriver

use of org.apache.kafka.streams.TopologyTestDriver in project kafka by apache.

the class ProcessorTopologyTest method testDrivingStatefulTopology.

@Test
public void testDrivingStatefulTopology() {
    final String storeName = "entries";
    driver = new TopologyTestDriver(createStatefulTopology(storeName), props);
    final TestInputTopic<String, String> inputTopic = driver.createInputTopic(INPUT_TOPIC_1, STRING_SERIALIZER, STRING_SERIALIZER);
    final TestOutputTopic<Integer, String> outputTopic1 = driver.createOutputTopic(OUTPUT_TOPIC_1, Serdes.Integer().deserializer(), Serdes.String().deserializer());
    inputTopic.pipeInput("key1", "value1");
    inputTopic.pipeInput("key2", "value2");
    inputTopic.pipeInput("key3", "value3");
    inputTopic.pipeInput("key1", "value4");
    assertTrue(outputTopic1.isEmpty());
    final KeyValueStore<String, String> store = driver.getKeyValueStore(storeName);
    assertEquals("value4", store.get("key1"));
    assertEquals("value2", store.get("key2"));
    assertEquals("value3", store.get("key3"));
    assertNull(store.get("key4"));
}
Also used : TopologyTestDriver(org.apache.kafka.streams.TopologyTestDriver) CoreMatchers.containsString(org.hamcrest.CoreMatchers.containsString) Test(org.junit.Test)

Example 49 with TopologyTestDriver

use of org.apache.kafka.streams.TopologyTestDriver in project kafka by apache.

the class RepartitionOptimizingTest method runTest.

private void runTest(final String optimizationConfig, final int expectedNumberRepartitionTopics) {
    final StreamsBuilder builder = new StreamsBuilder();
    final KStream<String, String> sourceStream = builder.stream(INPUT_TOPIC, Consumed.with(Serdes.String(), Serdes.String()).withName("sourceStream"));
    final KStream<String, String> mappedStream = sourceStream.map((k, v) -> KeyValue.pair(k.toUpperCase(Locale.getDefault()), v), Named.as("source-map"));
    mappedStream.filter((k, v) -> k.equals("B"), Named.as("process-filter")).mapValues(v -> v.toUpperCase(Locale.getDefault()), Named.as("process-mapValues")).process(() -> new SimpleProcessor(processorValueCollector), Named.as("process"));
    final KStream<String, Long> countStream = mappedStream.groupByKey(Grouped.as("count-groupByKey")).count(Named.as("count"), Materialized.<String, Long>as(Stores.inMemoryKeyValueStore("count-store")).withKeySerde(Serdes.String()).withValueSerde(Serdes.Long())).toStream(Named.as("count-toStream"));
    countStream.to(COUNT_TOPIC, Produced.with(Serdes.String(), Serdes.Long()).withName("count-to"));
    mappedStream.groupByKey(Grouped.as("aggregate-groupByKey")).aggregate(initializer, aggregator, Named.as("aggregate"), Materialized.<String, Integer>as(Stores.inMemoryKeyValueStore("aggregate-store")).withKeySerde(Serdes.String()).withValueSerde(Serdes.Integer())).toStream(Named.as("aggregate-toStream")).to(AGGREGATION_TOPIC, Produced.with(Serdes.String(), Serdes.Integer()).withName("reduce-to"));
    // adding operators for case where the repartition node is further downstream
    mappedStream.filter((k, v) -> true, Named.as("reduce-filter")).peek((k, v) -> System.out.println(k + ":" + v), Named.as("reduce-peek")).groupByKey(Grouped.as("reduce-groupByKey")).reduce(reducer, Named.as("reducer"), Materialized.as(Stores.inMemoryKeyValueStore("reduce-store"))).toStream(Named.as("reduce-toStream")).to(REDUCE_TOPIC, Produced.with(Serdes.String(), Serdes.String()));
    mappedStream.filter((k, v) -> k.equals("A"), Named.as("join-filter")).join(countStream, (v1, v2) -> v1 + ":" + v2.toString(), JoinWindows.of(ofMillis(5000)), StreamJoined.<String, String, Long>with(Stores.inMemoryWindowStore("join-store", ofDays(1), ofMillis(10000), true), Stores.inMemoryWindowStore("other-join-store", ofDays(1), ofMillis(10000), true)).withName("join").withKeySerde(Serdes.String()).withValueSerde(Serdes.String()).withOtherValueSerde(Serdes.Long())).to(JOINED_TOPIC, Produced.as("join-to"));
    streamsConfiguration.setProperty(StreamsConfig.TOPOLOGY_OPTIMIZATION_CONFIG, optimizationConfig);
    final Topology topology = builder.build(streamsConfiguration);
    topologyTestDriver = new TopologyTestDriver(topology, streamsConfiguration);
    final TestInputTopic<String, String> inputTopicA = topologyTestDriver.createInputTopic(INPUT_TOPIC, stringSerializer, stringSerializer);
    final TestOutputTopic<String, Long> countOutputTopic = topologyTestDriver.createOutputTopic(COUNT_TOPIC, stringDeserializer, new LongDeserializer());
    final TestOutputTopic<String, Integer> aggregationOutputTopic = topologyTestDriver.createOutputTopic(AGGREGATION_TOPIC, stringDeserializer, new IntegerDeserializer());
    final TestOutputTopic<String, String> reduceOutputTopic = topologyTestDriver.createOutputTopic(REDUCE_TOPIC, stringDeserializer, stringDeserializer);
    final TestOutputTopic<String, String> joinedOutputTopic = topologyTestDriver.createOutputTopic(JOINED_TOPIC, stringDeserializer, stringDeserializer);
    inputTopicA.pipeKeyValueList(getKeyValues());
    // Verify the topology
    final String topologyString = topology.describe().toString();
    if (optimizationConfig.equals(StreamsConfig.OPTIMIZE)) {
        assertEquals(EXPECTED_OPTIMIZED_TOPOLOGY, topologyString);
    } else {
        assertEquals(EXPECTED_UNOPTIMIZED_TOPOLOGY, topologyString);
    }
    // Verify the number of repartition topics
    assertEquals(expectedNumberRepartitionTopics, getCountOfRepartitionTopicsFound(topologyString));
    // Verify the values collected by the processor
    assertThat(3, equalTo(processorValueCollector.size()));
    assertThat(processorValueCollector, equalTo(expectedCollectedProcessorValues));
    // Verify the expected output
    assertThat(countOutputTopic.readKeyValuesToMap(), equalTo(keyValueListToMap(expectedCountKeyValues)));
    assertThat(aggregationOutputTopic.readKeyValuesToMap(), equalTo(keyValueListToMap(expectedAggKeyValues)));
    assertThat(reduceOutputTopic.readKeyValuesToMap(), equalTo(keyValueListToMap(expectedReduceKeyValues)));
    assertThat(joinedOutputTopic.readKeyValuesToMap(), equalTo(keyValueListToMap(expectedJoinKeyValues)));
}
Also used : StreamsConfig(org.apache.kafka.streams.StreamsConfig) Arrays(java.util.Arrays) Produced(org.apache.kafka.streams.kstream.Produced) CoreMatchers.equalTo(org.hamcrest.CoreMatchers.equalTo) Stores(org.apache.kafka.streams.state.Stores) LoggerFactory(org.slf4j.LoggerFactory) HashMap(java.util.HashMap) KStream(org.apache.kafka.streams.kstream.KStream) StreamJoined(org.apache.kafka.streams.kstream.StreamJoined) ArrayList(java.util.ArrayList) Initializer(org.apache.kafka.streams.kstream.Initializer) JoinWindows(org.apache.kafka.streams.kstream.JoinWindows) Matcher(java.util.regex.Matcher) StringDeserializer(org.apache.kafka.common.serialization.StringDeserializer) Locale(java.util.Locale) Map(java.util.Map) Named(org.apache.kafka.streams.kstream.Named) After(org.junit.After) Serdes(org.apache.kafka.common.serialization.Serdes) StringSerializer(org.apache.kafka.common.serialization.StringSerializer) Record(org.apache.kafka.streams.processor.api.Record) Deserializer(org.apache.kafka.common.serialization.Deserializer) Processor(org.apache.kafka.streams.processor.api.Processor) MatcherAssert.assertThat(org.hamcrest.MatcherAssert.assertThat) Aggregator(org.apache.kafka.streams.kstream.Aggregator) Before(org.junit.Before) Duration.ofDays(java.time.Duration.ofDays) TopologyTestDriver(org.apache.kafka.streams.TopologyTestDriver) StreamsBuilder(org.apache.kafka.streams.StreamsBuilder) TestOutputTopic(org.apache.kafka.streams.TestOutputTopic) Properties(java.util.Properties) Logger(org.slf4j.Logger) Consumed(org.apache.kafka.streams.kstream.Consumed) KeyValue(org.apache.kafka.streams.KeyValue) LongDeserializer(org.apache.kafka.common.serialization.LongDeserializer) Test(org.junit.Test) Grouped(org.apache.kafka.streams.kstream.Grouped) List(java.util.List) Serializer(org.apache.kafka.common.serialization.Serializer) Reducer(org.apache.kafka.streams.kstream.Reducer) Materialized(org.apache.kafka.streams.kstream.Materialized) IntegerDeserializer(org.apache.kafka.common.serialization.IntegerDeserializer) TestInputTopic(org.apache.kafka.streams.TestInputTopic) StreamsTestUtils(org.apache.kafka.test.StreamsTestUtils) Pattern(java.util.regex.Pattern) Duration.ofMillis(java.time.Duration.ofMillis) Topology(org.apache.kafka.streams.Topology) Assert.assertEquals(org.junit.Assert.assertEquals) IntegerDeserializer(org.apache.kafka.common.serialization.IntegerDeserializer) TopologyTestDriver(org.apache.kafka.streams.TopologyTestDriver) Topology(org.apache.kafka.streams.Topology) StreamsBuilder(org.apache.kafka.streams.StreamsBuilder) LongDeserializer(org.apache.kafka.common.serialization.LongDeserializer)

Example 50 with TopologyTestDriver

use of org.apache.kafka.streams.TopologyTestDriver in project kafka by apache.

the class RepartitionWithMergeOptimizingTest method runTest.

private void runTest(final String optimizationConfig, final int expectedNumberRepartitionTopics) {
    streamsConfiguration.setProperty(StreamsConfig.TOPOLOGY_OPTIMIZATION_CONFIG, optimizationConfig);
    final StreamsBuilder builder = new StreamsBuilder();
    final KStream<String, String> sourceAStream = builder.stream(INPUT_A_TOPIC, Consumed.with(Serdes.String(), Serdes.String()).withName("sourceAStream"));
    final KStream<String, String> sourceBStream = builder.stream(INPUT_B_TOPIC, Consumed.with(Serdes.String(), Serdes.String()).withName("sourceBStream"));
    final KStream<String, String> mappedAStream = sourceAStream.map((k, v) -> KeyValue.pair(v.split(":")[0], v), Named.as("mappedAStream"));
    final KStream<String, String> mappedBStream = sourceBStream.map((k, v) -> KeyValue.pair(v.split(":")[0], v), Named.as("mappedBStream"));
    final KStream<String, String> mergedStream = mappedAStream.merge(mappedBStream, Named.as("mergedStream"));
    mergedStream.groupByKey(Grouped.as("long-groupByKey")).count(Named.as("long-count"), Materialized.as(Stores.inMemoryKeyValueStore("long-store"))).toStream(Named.as("long-toStream")).to(COUNT_TOPIC, Produced.with(Serdes.String(), Serdes.Long()).withName("long-to"));
    mergedStream.groupByKey(Grouped.as("string-groupByKey")).count(Named.as("string-count"), Materialized.as(Stores.inMemoryKeyValueStore("string-store"))).toStream(Named.as("string-toStream")).mapValues(v -> v.toString(), Named.as("string-mapValues")).to(STRING_COUNT_TOPIC, Produced.with(Serdes.String(), Serdes.String()).withName("string-to"));
    final Topology topology = builder.build(streamsConfiguration);
    topologyTestDriver = new TopologyTestDriver(topology, streamsConfiguration);
    final TestInputTopic<String, String> inputTopicA = topologyTestDriver.createInputTopic(INPUT_A_TOPIC, stringSerializer, stringSerializer);
    final TestInputTopic<String, String> inputTopicB = topologyTestDriver.createInputTopic(INPUT_B_TOPIC, stringSerializer, stringSerializer);
    final TestOutputTopic<String, Long> countOutputTopic = topologyTestDriver.createOutputTopic(COUNT_TOPIC, stringDeserializer, new LongDeserializer());
    final TestOutputTopic<String, String> stringCountOutputTopic = topologyTestDriver.createOutputTopic(STRING_COUNT_TOPIC, stringDeserializer, stringDeserializer);
    inputTopicA.pipeKeyValueList(getKeyValues());
    inputTopicB.pipeKeyValueList(getKeyValues());
    final String topologyString = topology.describe().toString();
    // Verify the topology
    if (optimizationConfig.equals(StreamsConfig.OPTIMIZE)) {
        assertEquals(EXPECTED_OPTIMIZED_TOPOLOGY, topologyString);
    } else {
        assertEquals(EXPECTED_UNOPTIMIZED_TOPOLOGY, topologyString);
    }
    // Verify the number of repartition topics
    assertEquals(expectedNumberRepartitionTopics, getCountOfRepartitionTopicsFound(topologyString));
    // Verify the expected output
    assertThat(countOutputTopic.readKeyValuesToMap(), equalTo(keyValueListToMap(expectedCountKeyValues)));
    assertThat(stringCountOutputTopic.readKeyValuesToMap(), equalTo(keyValueListToMap(expectedStringCountKeyValues)));
}
Also used : StreamsBuilder(org.apache.kafka.streams.StreamsBuilder) StreamsConfig(org.apache.kafka.streams.StreamsConfig) Arrays(java.util.Arrays) Produced(org.apache.kafka.streams.kstream.Produced) CoreMatchers.equalTo(org.hamcrest.CoreMatchers.equalTo) Stores(org.apache.kafka.streams.state.Stores) LoggerFactory(org.slf4j.LoggerFactory) HashMap(java.util.HashMap) KStream(org.apache.kafka.streams.kstream.KStream) ArrayList(java.util.ArrayList) Matcher(java.util.regex.Matcher) StringDeserializer(org.apache.kafka.common.serialization.StringDeserializer) Map(java.util.Map) Named(org.apache.kafka.streams.kstream.Named) After(org.junit.After) Serdes(org.apache.kafka.common.serialization.Serdes) StringSerializer(org.apache.kafka.common.serialization.StringSerializer) Deserializer(org.apache.kafka.common.serialization.Deserializer) MatcherAssert.assertThat(org.hamcrest.MatcherAssert.assertThat) Before(org.junit.Before) TopologyTestDriver(org.apache.kafka.streams.TopologyTestDriver) StreamsBuilder(org.apache.kafka.streams.StreamsBuilder) TestOutputTopic(org.apache.kafka.streams.TestOutputTopic) Properties(java.util.Properties) Logger(org.slf4j.Logger) Consumed(org.apache.kafka.streams.kstream.Consumed) KeyValue(org.apache.kafka.streams.KeyValue) LongDeserializer(org.apache.kafka.common.serialization.LongDeserializer) Test(org.junit.Test) Grouped(org.apache.kafka.streams.kstream.Grouped) List(java.util.List) Serializer(org.apache.kafka.common.serialization.Serializer) Materialized(org.apache.kafka.streams.kstream.Materialized) TestInputTopic(org.apache.kafka.streams.TestInputTopic) StreamsTestUtils(org.apache.kafka.test.StreamsTestUtils) Pattern(java.util.regex.Pattern) Topology(org.apache.kafka.streams.Topology) Assert.assertEquals(org.junit.Assert.assertEquals) LongDeserializer(org.apache.kafka.common.serialization.LongDeserializer) TopologyTestDriver(org.apache.kafka.streams.TopologyTestDriver) Topology(org.apache.kafka.streams.Topology)

Aggregations

TopologyTestDriver (org.apache.kafka.streams.TopologyTestDriver)240 Test (org.junit.Test)209 StringSerializer (org.apache.kafka.common.serialization.StringSerializer)152 StreamsBuilder (org.apache.kafka.streams.StreamsBuilder)132 MockApiProcessorSupplier (org.apache.kafka.test.MockApiProcessorSupplier)91 IntegerSerializer (org.apache.kafka.common.serialization.IntegerSerializer)63 KeyValue (org.apache.kafka.streams.KeyValue)60 CoreMatchers.containsString (org.hamcrest.CoreMatchers.containsString)57 KeyValueTimestamp (org.apache.kafka.streams.KeyValueTimestamp)54 Windowed (org.apache.kafka.streams.kstream.Windowed)53 StringDeserializer (org.apache.kafka.common.serialization.StringDeserializer)51 KeyValueStore (org.apache.kafka.streams.state.KeyValueStore)48 Topology (org.apache.kafka.streams.Topology)47 Properties (java.util.Properties)41 TestInputTopic (org.apache.kafka.streams.TestInputTopic)32 Serdes (org.apache.kafka.common.serialization.Serdes)31 Consumed (org.apache.kafka.streams.kstream.Consumed)30 StreamsTestUtils (org.apache.kafka.test.StreamsTestUtils)24 Bytes (org.apache.kafka.common.utils.Bytes)23 TestRecord (org.apache.kafka.streams.test.TestRecord)23