How to merge many streams into one stream

Problem:

you have many Kafka topics with events in them, and you want to merge them all into a single topic.

Edit this page

Example use case:

Suppose that you have a set of Kafka topics representing songs being played of a particular genre. You might have a topic for rock songs, another for classical songs, and so forth. In this tutorial, we'll write a program that merges all of the song play events into a single topic.

Code example:

Try it

1
Initialize the project

To get started, make a new directory anywhere you’d like for this project:

mkdir merge-streams && cd merge-streams

2
Get Confluent Platform

Next, create the following docker-compose.yml file to obtain Confluent Platform:

---
version: '2'

services:
  zookeeper:
    image: confluentinc/cp-zookeeper:5.3.0
    hostname: zookeeper
    container_name: zookeeper
    ports:
      - "2181:2181"
    environment:
      ZOOKEEPER_CLIENT_PORT: 2181
      ZOOKEEPER_TICK_TIME: 2000

  broker:
    image: confluentinc/cp-enterprise-kafka:5.3.0
    hostname: broker
    container_name: broker
    depends_on:
      - zookeeper
    ports:
      - "29092:29092"
    environment:
      KAFKA_BROKER_ID: 1
      KAFKA_ZOOKEEPER_CONNECT: 'zookeeper:2181'
      KAFKA_LISTENER_SECURITY_PROTOCOL_MAP: PLAINTEXT:PLAINTEXT,PLAINTEXT_HOST:PLAINTEXT
      KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://broker:9092,PLAINTEXT_HOST://localhost:29092
      KAFKA_METRIC_REPORTERS: io.confluent.metrics.reporter.ConfluentMetricsReporter
      KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR: 1
      KAFKA_GROUP_INITIAL_REBALANCE_DELAY_MS: 0
      CONFLUENT_METRICS_REPORTER_BOOTSTRAP_SERVERS: broker:9092
      CONFLUENT_METRICS_REPORTER_ZOOKEEPER_CONNECT: zookeeper:2181
      CONFLUENT_METRICS_REPORTER_TOPIC_REPLICAS: 1
      CONFLUENT_METRICS_ENABLE: 'true'
      CONFLUENT_SUPPORT_CUSTOMER_ID: 'anonymous'

  schema-registry:
    image: confluentinc/cp-schema-registry:5.3.0
    hostname: schema-registry
    container_name: schema-registry
    depends_on:
      - zookeeper
      - broker
    ports:
      - "8081:8081"
    environment:
      SCHEMA_REGISTRY_HOST_NAME: schema-registry
      SCHEMA_REGISTRY_KAFKASTORE_CONNECTION_URL: 'zookeeper:2181'
      SCHEMA_REGISTRY_LOG4J_ROOT_LOGLEVEL: WARN

And launch it by running:

docker-compose up

3
Configure the project

Create the following Gradle build file, named build.gradle for the project:

buildscript {
    repositories {
        jcenter()
    }
    dependencies {
        classpath "com.commercehub.gradle.plugin:gradle-avro-plugin:0.15.1"
        classpath "com.github.jengelman.gradle.plugins:shadow:4.0.2"
    }
}

plugins {
    id "java"
    id "com.google.cloud.tools.jib" version "1.1.1"
}

sourceCompatibility = "1.8"
targetCompatibility = "1.8"
version = "0.0.1"

repositories {
    mavenCentral()
    jcenter()

    maven {
        url "http://packages.confluent.io/maven"
    }
}

apply plugin: "com.commercehub.gradle.plugin.avro"
apply plugin: "com.github.johnrengelman.shadow"

dependencies {
    compile "org.apache.avro:avro:1.8.2"
    implementation "org.slf4j:slf4j-simple:1.7.26"
    implementation "org.apache.kafka:kafka-streams:2.2.0"
    implementation "io.confluent:kafka-streams-avro-serde:5.2.0"
    testCompile "org.apache.kafka:kafka-streams-test-utils:2.2.0"
    testCompile "junit:junit:4.12"
}

test {
    testLogging {
        outputs.upToDateWhen { false }
        showStandardStreams = true
        exceptionFormat = "full"
    }
}

jar {
  manifest {
    attributes(
      "Class-Path": configurations.compile.collect { it.getName() }.join(" "),
      "Main-Class": "io.confluent.developer.MergeStreams"
    )
  }
}

shadowJar {
    archiveName = "kstreams-merge-standalone-${version}.${extension}"
}

And be sure to run the following command to obtain the Gradle wrapper:

gradle wrapper

Next, create a directory for configuration data:

mkdir configuration

Then create a development file at configuration/dev.properties:

application.id=merging-app
bootstrap.servers=127.0.0.1:29092
schema.registry.url=http://127.0.0.1:8081

input.rock.topic.name=rock-song-events
input.rock.topic.partitions=1
input.rock.topic.replication.factor=1

input.classical.topic.name=classical-song-events
input.classical.topic.partitions=1
input.classical.topic.replication.factor=1

output.topic.name=all-song-events
output.topic.partitions=1
output.topic.replication.factor=1

4
Create a schema for the events

Create a directory for the schemas that represent the events in the stream:

mkdir -p src/main/avro

Then create the following Avro schema file at src/main/avro/song_event.avsc for the events representing a song being played:

{
  "namespace": "io.confluent.developer.avro",
  "type": "record",
  "name": "SongEvent",
  "fields": [
    {"name": "artist", "type": "string"},
    {"name": "title", "type": "string"}
  ]
}

Because we will use this Avro schema in our Java code, we’ll need to compile it. Run the following:

./gradlew build

5
Create the Kafka Streams topology

Create a directory for the Java files in this project:

mkdir -p src/main/java/io/confluent/developer

Then create the following file at src/main/java/io/confluent/developer/MergeStreams.java. Notice the buildTopology method, which uses the Kafka Streams DSL. A stream is opened up for each input topic. The input streams are then combined using the merge function, which creates a new stream that represents all of the events of its inputs. Note that you can chain merge to combine as many streams as needed. The merged stream is then connected to the to method, which the name of a Kafka topic to send the events to.

package io.confluent.developer;

import io.confluent.developer.avro.SongEvent;
import io.confluent.kafka.serializers.AbstractKafkaAvroSerDeConfig;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroSerde;
import org.apache.kafka.clients.admin.AdminClient;
import org.apache.kafka.clients.admin.NewTopic;
import org.apache.kafka.common.serialization.Serdes;
import org.apache.kafka.streams.KafkaStreams;
import org.apache.kafka.streams.StreamsBuilder;
import org.apache.kafka.streams.StreamsConfig;
import org.apache.kafka.streams.Topology;
import org.apache.kafka.streams.kstream.KStream;

import java.io.FileInputStream;
import java.io.IOException;
import java.util.*;
import java.util.concurrent.CountDownLatch;

public class MergeStreams {

    public Properties buildStreamsProperties(Properties envProps) {
        Properties props = new Properties();

        props.put(StreamsConfig.APPLICATION_ID_CONFIG, envProps.getProperty("application.id"));
        props.put(StreamsConfig.BOOTSTRAP_SERVERS_CONFIG, envProps.getProperty("bootstrap.servers"));
        props.put(StreamsConfig.DEFAULT_KEY_SERDE_CLASS_CONFIG, Serdes.String().getClass());
        props.put(StreamsConfig.DEFAULT_VALUE_SERDE_CLASS_CONFIG, SpecificAvroSerde.class);
        props.put(AbstractKafkaAvroSerDeConfig.SCHEMA_REGISTRY_URL_CONFIG, envProps.getProperty("schema.registry.url"));

        return props;
    }

    public Topology buildTopology(Properties envProps) {
        final StreamsBuilder builder = new StreamsBuilder();

        final String rockTopic = envProps.getProperty("input.rock.topic.name");
        final String classicalTopic = envProps.getProperty("input.classical.topic.name");
        final String allGenresTopic = envProps.getProperty("output.topic.name");

        KStream<String, SongEvent> rockSongs = builder.stream(rockTopic);
        KStream<String, SongEvent> classicalSongs = builder.stream(classicalTopic);
        KStream<String, SongEvent> allSongs = rockSongs.merge(classicalSongs);

        allSongs.to(allGenresTopic);
        return builder.build();
    }

    public void createTopics(Properties envProps) {
        Map<String, Object> config = new HashMap<>();
        config.put("bootstrap.servers", envProps.getProperty("bootstrap.servers"));
        AdminClient client = AdminClient.create(config);

        List<NewTopic> topics = new ArrayList<>();

        topics.add(new NewTopic(
                envProps.getProperty("input.rock.topic.name"),
                Integer.parseInt(envProps.getProperty("input.rock.topic.partitions")),
                Short.parseShort(envProps.getProperty("input.rock.topic.replication.factor"))));

        topics.add(new NewTopic(
                envProps.getProperty("input.classical.topic.name"),
                Integer.parseInt(envProps.getProperty("input.classical.topic.partitions")),
                Short.parseShort(envProps.getProperty("input.classical.topic.replication.factor"))));

        topics.add(new NewTopic(
                envProps.getProperty("output.topic.name"),
                Integer.parseInt(envProps.getProperty("output.topic.partitions")),
                Short.parseShort(envProps.getProperty("output.topic.replication.factor"))));

        client.createTopics(topics);
        client.close();
    }

    public Properties loadEnvProperties(String fileName) throws IOException {
        Properties envProps = new Properties();
        FileInputStream input = new FileInputStream(fileName);
        envProps.load(input);
        input.close();

        return envProps;
    }

    public static void main(String[] args) throws Exception {
        if (args.length < 1) {
            throw new IllegalArgumentException("This program takes one argument: the path to an environment configuration file.");
        }

        MergeStreams ms = new MergeStreams();
        Properties envProps = ms.loadEnvProperties(args[0]);
        Properties streamProps = ms.buildStreamsProperties(envProps);
        Topology topology = ms.buildTopology(envProps);

        ms.createTopics(envProps);

        final KafkaStreams streams = new KafkaStreams(topology, streamProps);
        final CountDownLatch latch = new CountDownLatch(1);

        // Attach shutdown handler to catch Control-C.
        Runtime.getRuntime().addShutdownHook(new Thread("streams-shutdown-hook") {
            @Override
            public void run() {
                streams.close();
                latch.countDown();
            }
        });

        try {
            streams.start();
            latch.await();
        } catch (Throwable e) {
            System.exit(1);
        }
        System.exit(0);
    }
}

6
Compile and run the Kafka Streams program

In your terminal, run:

./gradlew shadowJar

Now that an uberjar for the Kafka Streams application has been built, you can launch it locally. When you run the following, the prompt won’t return, because the application will run until you exit it:

java -jar build/libs/kstreams-merge-standalone-0.0.1.jar configuration/dev.properties

7
Produce events to the input topics

To produce the input events to their respective topics, you’ll want two terminals running. To send the rock songs to their topic, open up a terminal and run the following:

docker exec -i schema-registry /usr/bin/kafka-avro-console-producer --topic rock-song-events --broker-list broker:9092 --property value.schema="$(< src/main/avro/song_event.avsc)"

When the console producer starts, it will log some messages and hang, waiting for your input. Type in one line at a time and press enter to send it. Each line represents an event. To send all of the events below, paste the following into the prompt and press enter:

{"artist": "Metallica", "title": "Fade to Black"}
{"artist": "Smashing Pumpkins", "title": "Today"}
{"artist": "Pink Floyd", "title": "Another Brick in the Wall"}
{"artist": "Van Halen", "title": "Jump"}
{"artist": "Led Zeppelin", "title": "Kashmir"}

To produce the classical songs, open up another terminal and run:

docker exec -i schema-registry /usr/bin/kafka-avro-console-producer --topic classical-song-events --broker-list broker:9092 --property value.schema="$(< src/main/avro/song_event.avsc)"

Then paste in the following events:

{"artist": "Wolfgang Amadeus Mozart", "title": "The Magic Flute"}
{"artist": "Johann Pachelbel", "title": "Canon"}
{"artist": "Ludwig van Beethoven", "title": "Symphony No. 5"}
{"artist": "Edward Elgar", "title": "Pomp and Circumstance"}

8
Consume events from the output topic

Leaving your original terminals running, open another to consume the events that have been merged:

docker exec -it schema-registry /usr/bin/kafka-avro-console-consumer --topic all-song-events --bootstrap-server broker:9092 --from-beginning

After the consumer starts, you should see the following messages. The order might vary depending on the timing of which the input events are sent to each topic and processed by the app. Kafka Streams will coalesce the respective input topics together in an indeterminate manner. To continue studying the example, send more events through the input terminal prompt. Otherwise, you can Control-C to exit the process.

{"artist":"Metallica","title":"Fade to Black"}
{"artist":"Smashing Pumpkins","title":"Today"}
{"artist":"Pink Floyd","title":"Another Brick in the Wall"}
{"artist":"Van Halen","title":"Jump"}
{"artist":"Led Zeppelin","title":"Kashmir"}
{"artist":"Wolfgang Amadeus Mozart","title":"The Magic Flute"}
{"artist":"Johann Pachelbel","title":"Canon"}
{"artist":"Ludwig van Beethoven","title":"Symphony No. 5"}
{"artist":"Edward Elgar","title":"Pomp and Circumstance"}

Test it

1
Create a test configuration file

First, create a test file at configuration/test.properties:

application.id=merging-app
bootstrap.servers=127.0.0.1:29092
schema.registry.url=http://127.0.0.1:8081

input.rock.topic.name=rock-song-events
input.rock.topic.partitions=1
input.rock.topic.replication.factor=1

input.classical.topic.name=classical-song-events
input.classical.topic.partitions=1
input.classical.topic.replication.factor=1

output.topic.name=all-song-events
output.topic.partitions=1
output.topic.replication.factor=1

2
Write a test

Then, create a directory for the tests to live in:

mkdir -p src/test/java/io/confluent/developer

Create the following test file at src/test/java/io/confluent/developer/MergeStreamsTest.java:

package io.confluent.developer;

import io.confluent.developer.avro.SongEvent;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroDeserializer;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroSerializer;
import org.apache.kafka.clients.producer.ProducerRecord;
import org.apache.kafka.common.serialization.Deserializer;
import org.apache.kafka.common.serialization.Serdes;
import org.apache.kafka.common.serialization.Serializer;
import org.apache.kafka.streams.Topology;
import org.apache.kafka.streams.TopologyTestDriver;
import org.apache.kafka.streams.test.ConsumerRecordFactory;
import org.junit.Assert;
import org.junit.Test;

import java.io.IOException;
import java.util.*;

public class MergeStreamsTest {

    private final static String TEST_CONFIG_FILE = "configuration/test.properties";

    public SpecificAvroSerializer<SongEvent> makeSerializer(Properties envProps) {
        SpecificAvroSerializer<SongEvent> serializer = new SpecificAvroSerializer<>();

        Map<String, String> config = new HashMap<>();
        config.put("schema.registry.url", envProps.getProperty("schema.registry.url"));
        serializer.configure(config, false);

        return serializer;
    }

    public SpecificAvroDeserializer<SongEvent> makeDeserializer(Properties envProps) {
        SpecificAvroDeserializer<SongEvent> deserializer = new SpecificAvroDeserializer<>();

        Map<String, String> config = new HashMap<>();
        config.put("schema.registry.url", envProps.getProperty("schema.registry.url"));
        deserializer.configure(config, false);

        return deserializer;
    }

    @Test
    public void testMergeStreams() throws IOException {
        MergeStreams ms = new MergeStreams();
        Properties envProps = ms.loadEnvProperties(TEST_CONFIG_FILE);
        Properties streamProps = ms.buildStreamsProperties(envProps);

        String rockTopic = envProps.getProperty("input.rock.topic.name");
        String classicalTopic = envProps.getProperty("input.classical.topic.name");
        String allGenresTopic = envProps.getProperty("output.topic.name");

        Topology topology = ms.buildTopology(envProps);
        TopologyTestDriver testDriver = new TopologyTestDriver(topology, streamProps);

        Serializer<String> keySerializer = Serdes.String().serializer();
        SpecificAvroSerializer<SongEvent> valueSerializer = makeSerializer(envProps);

        Deserializer<String> keyDeserializer = Serdes.String().deserializer();
        SpecificAvroDeserializer<SongEvent> valueDeserializer = makeDeserializer(envProps);

        ConsumerRecordFactory<String, SongEvent> inputFactory = new ConsumerRecordFactory<>(keySerializer, valueSerializer);

        List<SongEvent> rockSongs = new ArrayList<>();
        List<SongEvent> classicalSongs = new ArrayList<>();

        rockSongs.add(SongEvent.newBuilder().setArtist("Metallica").setTitle("Fade to Black").build());
        rockSongs.add(SongEvent.newBuilder().setArtist("Smashing Pumpkins").setTitle("Today").build());
        rockSongs.add(SongEvent.newBuilder().setArtist("Pink Floyd").setTitle("Another Brick in the Wall").build());
        rockSongs.add(SongEvent.newBuilder().setArtist("Van Halen").setTitle("Jump").build());
        rockSongs.add(SongEvent.newBuilder().setArtist("Led Zeppelin").setTitle("Kashmir").build());

        classicalSongs.add(SongEvent.newBuilder().setArtist("Wolfgang Amadeus Mozart").setTitle("The Magic Flute").build());
        classicalSongs.add(SongEvent.newBuilder().setArtist("Johann Pachelbel").setTitle("Canon").build());
        classicalSongs.add(SongEvent.newBuilder().setArtist("Ludwig van Beethoven").setTitle("Symphony No. 5").build());
        classicalSongs.add(SongEvent.newBuilder().setArtist("Edward Elgar").setTitle("Pomp and Circumstance").build());

        for (SongEvent song : rockSongs) {
            testDriver.pipeInput(inputFactory.create(rockTopic, song.getArtist(), song));
        }

        for (SongEvent song : classicalSongs) {
            testDriver.pipeInput(inputFactory.create(classicalTopic, song.getArtist(), song));
        }

        List<SongEvent> actualOutput = new ArrayList<>();
        while (true) {
            ProducerRecord<String, SongEvent> record = testDriver.readOutput(allGenresTopic, keyDeserializer, valueDeserializer);

            if (record != null) {
                actualOutput.add(record.value());
            } else {
                break;
            }
        }

        List<SongEvent> expectedOutput = new ArrayList<>();
        expectedOutput.addAll(rockSongs);
        expectedOutput.addAll(classicalSongs);

        Assert.assertEquals(expectedOutput, actualOutput);
    }

}

3
Invoke the tests

Now run the test, which is as simple as:

./gradlew test

Take it to production

1
Create a production configuration file

First, create a new configuration file at configuration/prod.properties with the following content. Be sure to fill in the addresses of your production hosts and change any other parameters that make sense for your setup.

application.id=merging-app
bootstrap.servers=<< FILL ME IN >>
schema.registry.url=<< FILL ME IN >>

input.rock.topic.name=rock-song-events
input.rock.topic.partitions=<< FILL ME IN >>
input.rock.topic.replication.factor=<< FILL ME IN >>

input.classical.topic.name=classical-song-events
input.classical.topic.partitions=<< FILL ME IN >>
input.classical.topic.replication.factor=<< FILL ME IN >>

output.topic.name=all-song-events
output.topic.partitions=<< FILL ME IN >>
output.topic.replication.factor=<< FILL ME IN >>

2
Build a Docker image

In your terminal, execute the following to invoke the Jib plugin to build an image:

./gradlew jibDockerBuild --image=io.confluent.developer/kstreams-merge:0.0.1

3
Launch the container

Finally, launch the container using your preferred container orchestration service. If you want to run it locally, you can execute the following:

docker run -v $PWD/configuration/prod.properties:/config.properties io.confluent.developer/kstreams-merge:0.0.1 config.properties