How to merge many streams into one stream

Question:

If you have many Kafka topics with events, how do you merge them all into a single topic?

Example use case:

Suppose that you have a set of Kafka topics representing songs of a particular genre being played. 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. Related pattern: Event Stream Merger.

Hands-on code example:

Confluent Cloud ksqlDB Kafka Streams Basic Kafka

Short Answer

The input streams are combined using the merge function, which creates a new stream that represents all of the events of its inputs. The merged stream is forwarded to a combined topic via the to method, which accepts the topic as a parameter.

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

allSongs.to(allGenresTopic);

Run it

Prerequisites

This tutorial installs Confluent Platform using Docker. Before proceeding:

• Install Docker Desktop (version 4.0.0 or later) or Docker Engine (version 19.03.0 or later) if you don’t already have it
• Install the Docker Compose plugin if you don’t already have it. This isn’t necessary if you have Docker Desktop since it includes Docker Compose.
• Start Docker if it’s not already running, either by starting Docker Desktop or, if you manage Docker Engine with systemd, via systemctl
• Verify that Docker is set up properly by ensuring no errors are output when you run docker info and docker compose version on the command line

Initialize the project

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

mkdir merge-streams && cd merge-streams

Get Confluent Platform

Next, create the following docker-compose.yml file to obtain Confluent Platform (for Kafka in the cloud, see Confluent Cloud):

version: '2'
services:
  broker:
    image: confluentinc/cp-kafka:7.4.1
    hostname: broker
    container_name: broker
    ports:
    - 29092:29092
    environment:
      KAFKA_BROKER_ID: 1
      KAFKA_LISTENER_SECURITY_PROTOCOL_MAP: PLAINTEXT:PLAINTEXT,PLAINTEXT_HOST:PLAINTEXT,CONTROLLER:PLAINTEXT
      KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://broker:9092,PLAINTEXT_HOST://localhost:29092
      KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR: 1
      KAFKA_GROUP_INITIAL_REBALANCE_DELAY_MS: 0
      KAFKA_TRANSACTION_STATE_LOG_MIN_ISR: 1
      KAFKA_TRANSACTION_STATE_LOG_REPLICATION_FACTOR: 1
      KAFKA_PROCESS_ROLES: broker,controller
      KAFKA_NODE_ID: 1
      KAFKA_CONTROLLER_QUORUM_VOTERS: 1@broker:29093
      KAFKA_LISTENERS: PLAINTEXT://broker:9092,CONTROLLER://broker:29093,PLAINTEXT_HOST://0.0.0.0:29092
      KAFKA_INTER_BROKER_LISTENER_NAME: PLAINTEXT
      KAFKA_CONTROLLER_LISTENER_NAMES: CONTROLLER
      KAFKA_LOG_DIRS: /tmp/kraft-combined-logs
      CLUSTER_ID: MkU3OEVBNTcwNTJENDM2Qk
  schema-registry:
    image: confluentinc/cp-schema-registry:7.3.0
    hostname: schema-registry
    container_name: schema-registry
    depends_on:
    - broker
    ports:
    - 8081:8081
    environment:
      SCHEMA_REGISTRY_HOST_NAME: schema-registry
      SCHEMA_REGISTRY_KAFKASTORE_BOOTSTRAP_SERVERS: broker:9092
      SCHEMA_REGISTRY_LOG4J_ROOT_LOGLEVEL: WARN

And launch it by running:

docker compose up -d

Configure the project

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

buildscript {
    repositories {
        mavenCentral()
    }
    dependencies {
        classpath "gradle.plugin.com.github.jengelman.gradle.plugins:shadow:7.0.0"
    }
}

plugins {
    id "java"
    id "com.github.davidmc24.gradle.plugin.avro" version "1.7.0"
}

sourceCompatibility = JavaVersion.VERSION_17
targetCompatibility = JavaVersion.VERSION_17
version = "0.0.1"

repositories {
    mavenCentral()

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

apply plugin: "com.github.johnrengelman.shadow"

dependencies {
    implementation "org.apache.avro:avro:1.11.1"
    implementation "org.slf4j:slf4j-simple:2.0.7"
    implementation 'org.apache.kafka:kafka-streams:3.4.0'
    implementation ('org.apache.kafka:kafka-clients') {
       version {
           strictly '3.4.0'
        }
      }
    implementation "io.confluent:kafka-streams-avro-serde:7.3.0"
    testImplementation "org.apache.kafka:kafka-streams-test-utils:3.4.0"
    testImplementation "junit:junit:4.13.2"
}

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

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

shadowJar {
    archiveBaseName = "kstreams-merge-standalone"
    archiveClassifier = ''
}

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

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

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 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.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.time.Duration;
import java.util.Properties;
import java.util.concurrent.CountDownLatch;

import io.confluent.developer.avro.SongEvent;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroSerde;

import static io.confluent.kafka.serializers.AbstractKafkaSchemaSerDeConfig.SCHEMA_REGISTRY_URL_CONFIG;

public class MergeStreams {

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

        final String rockTopic = allProps.getProperty("input.rock.topic.name");
        final String classicalTopic = allProps.getProperty("input.classical.topic.name");
        final String allGenresTopic = allProps.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 allProps) {
        AdminClient client = AdminClient.create(allProps);

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

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

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

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

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

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

        return allProps;
    }

    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 allProps = ms.loadEnvProperties(args[0]);
        allProps.put(StreamsConfig.DEFAULT_KEY_SERDE_CLASS_CONFIG, Serdes.String().getClass());
        allProps.put(StreamsConfig.DEFAULT_VALUE_SERDE_CLASS_CONFIG, SpecificAvroSerde.class);
        allProps.put(SCHEMA_REGISTRY_URL_CONFIG, allProps.getProperty("schema.registry.url"));
        Topology topology = ms.buildTopology(allProps);

        ms.createTopics(allProps);

        final KafkaStreams streams = new KafkaStreams(topology, allProps);
        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(Duration.ofSeconds(5));
                latch.countDown();
            }
        });

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

Note when using the merge operator the keys and values of the two KStream objects you’re merging must be of the same type. If you have 2 KStream instances with different key and/or value types, you’ll have to use the KStream.map (or KStream.mapValues) operation first to get the types to line-up before merging.

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

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 --bootstrap-server 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 --bootstrap-server 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"}

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

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=mock://merging-app: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

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 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.TestInputTopic;
import org.apache.kafka.streams.Topology;
import org.apache.kafka.streams.TopologyTestDriver;
import org.apache.kafka.streams.StreamsConfig;
import org.junit.After;
import org.junit.Assert;
import org.junit.Test;

import java.io.IOException;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Properties;
import java.util.stream.Collectors;

import io.confluent.developer.avro.SongEvent;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroDeserializer;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroSerializer;

public class MergeStreamsTest {

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

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

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

        return serializer;
    }

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

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

        return deserializer;
    }

    @Test
    public void testMergeStreams() throws IOException {
        MergeStreams ms = new MergeStreams();
        Properties allProps = ms.loadEnvProperties(TEST_CONFIG_FILE);
        allProps.put(StreamsConfig.DEFAULT_KEY_SERDE_CLASS_CONFIG, Serdes.String().getClass().getName());
        allProps.put(StreamsConfig.DEFAULT_VALUE_SERDE_CLASS_CONFIG, Serdes.String().getClass().getName());

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

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

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

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

        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());

        final TestInputTopic<String, SongEvent>
            rockSongsTestDriverTopic =
            testDriver.createInputTopic(rockTopic, keySerializer, valueSerializer);

        final TestInputTopic<String, SongEvent>
            classicRockSongsTestDriverTopic =
            testDriver.createInputTopic(classicalTopic, keySerializer, valueSerializer);

        for (SongEvent song : rockSongs) {
            rockSongsTestDriverTopic.pipeInput(song.getArtist(), song);
        }

        for (SongEvent song : classicalSongs) {
            classicRockSongsTestDriverTopic.pipeInput(song.getArtist(), song);
        }

        List<SongEvent> actualOutput =
            testDriver
                .createOutputTopic(allGenresTopic, keyDeserializer, valueDeserializer)
                .readKeyValuesToList()
                .stream()
                .filter(record -> record.value != null)
                .map(record -> record.value)
                .collect(Collectors.toList());

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

        Assert.assertEquals(expectedOutput, actualOutput);
    }

    @After
    public void cleanup() {
        testDriver.close();
    }
}

Invoke the tests

Now run the test, which is as simple as:

./gradlew test

Deploy on Confluent Cloud

Run your app with Confluent Cloud

Instead of running a local Kafka cluster, you may use Confluent Cloud, a fully managed Apache Kafka service.

Sign up for Confluent Cloud, a fully managed Apache Kafka service.
After you log in to Confluent Cloud Console, click Environments in the lefthand navigation, click on Add cloud environment, and name the environment learn-kafka. Using a new environment keeps your learning resources separate from your other Confluent Cloud resources.
From the Billing & payment section in the menu, apply the promo code CC100KTS to receive an additional $100 free usage on Confluent Cloud (details). To avoid having to enter a credit card, add an additional promo code CONFLUENTDEV1. With this promo code, you will not have to enter a credit card for 30 days or until your credits run out.
Click on LEARN and follow the instructions to launch a Kafka cluster and enable Schema Registry.

Next, from the Confluent Cloud Console, click on Clients to get the cluster-specific configurations, e.g., Kafka cluster bootstrap servers and credentials, Confluent Cloud Schema Registry and credentials, etc., and set the appropriate parameters in your client application. In the case of this tutorial, add the following properties to the client application’s input properties file, substituting all curly braces with your Confluent Cloud values.

# Required connection configs for Kafka producer, consumer, and admin
bootstrap.servers={{ BROKER_ENDPOINT }}
security.protocol=SASL_SSL
sasl.jaas.config=org.apache.kafka.common.security.plain.PlainLoginModule required username='{{ CLUSTER_API_KEY }}' password='{{ CLUSTER_API_SECRET }}';
sasl.mechanism=PLAIN
# Required for correctness in Apache Kafka clients prior to 2.6
client.dns.lookup=use_all_dns_ips

# Best practice for Kafka producer to prevent data loss
acks=all

# Required connection configs for Confluent Cloud Schema Registry
schema.registry.url=https://{{ SR_ENDPOINT }}
basic.auth.credentials.source=USER_INFO
schema.registry.basic.auth.user.info={{ SR_API_KEY }}:{{ SR_API_SECRET }}

Now you’re all set to run your streaming application locally, backed by a Kafka cluster fully managed by Confluent Cloud.