How build your first Apache KafkaProducer application

Question:

How do I get started in building my first Kafka producer application?

Edit this page

Example use case:

You'd like to integrate an Apache KafkaProducer in your event-driven application, but you're not sure where to start. In this tutorial you'll build a small application writing records to Kafka with a KafkaProducer. You can use the code in this tutorial as an example of how to use an Apache Kafka producer

Code example:

Try it

1
Initialize the project

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

mkdir kafka-producer-application && cd kafka-producer-application

2
Get Confluent Platform

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

---
version: '2'

services:
  zookeeper:
    image: confluentinc/cp-zookeeper:5.5.0
    hostname: zookeeper
    container_name: zookeeper
    ports:
      - "2181:2181"
    environment:
      ZOOKEEPER_CLIENT_PORT: 2181
      ZOOKEEPER_TICK_TIME: 2000

  broker:
    image: confluentinc/cp-kafka:5.5.0
    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_OFFSETS_TOPIC_REPLICATION_FACTOR: 1
      KAFKA_GROUP_INITIAL_REBALANCE_DELAY_MS: 0
      KAFKA_TOOLS_LOG4J_LOGLEVEL: ERROR


  schema-registry:
    image: confluentinc/cp-schema-registry:5.5.0
    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 -d

3
Create a topic

In this step we’re going to create a topic for use during this tutorial.

Open a new terminal window and then run this command to open a shell on the broker docker container

docker-compose exec broker bash

Next, create the topic that the producer to write to

kafka-topics --create --topic output-topic --bootstrap-server broker:9092 --replication-factor 1 --partitions 1

Keep this terminal window open for later use when you run a console consumer to verify your producer application.

4
Configure the project

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

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"
    id "idea"
    id "eclipse"
}

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

repositories {
    jcenter()

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

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

dependencies {
    implementation "org.apache.avro:avro:1.9.1"
    implementation "org.slf4j:slf4j-simple:1.7.30"
    implementation "org.apache.kafka:kafka-streams:2.5.0"
    implementation "io.confluent:kafka-streams-avro-serde:5.5.0"
    testImplementation "org.apache.kafka:kafka-streams-test-utils:2.5.0"
    testImplementation "junit:junit:4.12"
    testImplementation 'org.hamcrest:hamcrest:2.2'
}

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.KafkaProducerApplication"
    )
  }
}

shadowJar {
    archiveBaseName = "kafka-producer-application-standalone"
    archiveClassifier = ''
}

Run the following command to obtain the Gradle wrapper:

gradle wrapper

Next, create a directory for configuration data:

mkdir configuration

5
Add application and producer properties

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

bootstrap.servers=localhost:29092

key.serializer=org.apache.kafka.common.serialization.StringSerializer
value.serializer=org.apache.kafka.common.serialization.StringSerializer
acks=all

#Properties below this line are specific to code in this application
input.topic.name=input-topic
output.topic.name=output-topic

Let’s do a quick walkthrough of some of the producer properties.

key.serializer - The serializer the KafkaProducer will use to serialize the key.

value.serializer - The serializer the KafkaProducer will use to serialize the value.

acks - The KafkaProducer uses the acks configuration to tell the lead broker how many acknowledements to wait for to consider a produce request complete. Acceptable values for acks are: 0, 1 (the default), -1, or all. Setting acks to -1 is the same as setting it to all.

  • acks=0: "fire and forget", once the producer sends the record batch it is considered successful

  • acks=1: leader broker added the records to its local log but didn’t wait for any acknowledgement from the followers

  • acks=all: highest data durability guarantee, the leader broker persisted the record to its log and received acknowledgement of replication from all in-sync replicas When using aks=all, it’s stongly reccomended to update min.insync.replicas as well.

This is only a small sub-set of producer configuration parameters. The full list of producer configuration parameters can be found in the Apache Kafka documentation.

6
Create the KafkaProducer application

Create a directory for the Java files in this project:

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

Before you create your application file, let’s look at some of the key points of this program:

KafkaProducerApplication constructor
public class KafkaProducerApplication {

  private final Producer<String, String> producer;
  final String outTopic;

  public KafkaProducerApplication(final Producer<String, String> producer,  (1)
                                  final String topic) {                     (2)
    this.producer = producer;
    outTopic = topic;
  }
1 Passing in the Producer instance as a constructor parameter.
2 The topic to write records to

In this tutorial you’ll inject the dependencies in the KafkaProducerApplication.main() method. Having this thin wrapper class around a Producer is not required, but it does help with making our code easier to test. We’ll go into more details in the testing section of the tutorial.

(In practice you may want to use a dependency injection framework library, such as the Spring Framework).

Next let’s take a look at the KafkaProducerApplication.produce method

KafkaProducerApplication.produce
public Future<RecordMetadata> produce(final String message) {
    final String[] parts = message.split("-");  (1)
    final String key, value;
    if (parts.length > 1) {
      key = parts[0];
      value = parts[1];
    } else {
      key = "NO-KEY";
      value = parts[0];
    }
    final ProducerRecord<String, String> producerRecord = new ProducerRecord<>(outTopic, key, value);  (2)
    return producer.send(producerRecord);                 (3)
  }
1 Process the String for sending message
2 Create the ProducerRecord
3 Send the record to the broker

The KafkaProducerApplication.produce method does some processing on a String, and then sends the ProducerRecord. While this code is a trivial example, it’s enough to show the example of using a KafkaProducer. Notice that KafkaProducer.send returns a Future with a type of RecordMetadata.

The KafkaProducer.send method is asynchronous and returns as soon as the provided record is placed in the buffer of records to be sent to the broker. Once the broker acknowledges that the record has been appended to its log, the broker completes the produce request, which the application receives as RecordMetadata—information about the committed message. This tutorial prints the timestamp and offset for each record sent using the RecordMetadata object. Note that calling Future.get() for any record will block until the produce request completes.

Now go ahead and create the following file at src/main/java/io/confluent/developer/KafkaProducerApplication.java.

package io.confluent.developer;


import org.apache.kafka.clients.producer.KafkaProducer;
import org.apache.kafka.clients.producer.Producer;
import org.apache.kafka.clients.producer.ProducerRecord;
import org.apache.kafka.clients.producer.RecordMetadata;

import java.io.FileInputStream;
import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Paths;
import java.util.Collection;
import java.util.List;
import java.util.Properties;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.Future;
import java.util.stream.Collectors;

public class KafkaProducerApplication {

    private final Producer<String, String> producer;
    final String outTopic;

    public KafkaProducerApplication(final Producer<String, String> producer,
                                    final String topic) {
        this.producer = producer;
        outTopic = topic;
    }

    public Future<RecordMetadata> produce(final String message) {
        final String[] parts = message.split("-");
        final String key, value;
        if (parts.length > 1) {
            key = parts[0];
            value = parts[1];
        } else {
            key = "NO-KEY";
            value = parts[0];
        }
        final ProducerRecord<String, String> producerRecord = new ProducerRecord<>(outTopic, key, value);
        return producer.send(producerRecord);
    }

    public void shutdown() {
        producer.close();
    }

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

        return envProps;
    }

    public void printMetadata(final Collection<Future<RecordMetadata>> metadata,
                              final String fileName) {
        System.out.println("Offsets and timestamps committed in batch from " + fileName);
        metadata.forEach(m -> {
            try {
                final RecordMetadata recordMetadata = m.get();
                System.out.println("Record written to offset " + recordMetadata.offset() + " timestamp " + recordMetadata.timestamp());
            } catch (InterruptedException | ExecutionException e) {
                if (e instanceof InterruptedException) {
                    Thread.currentThread().interrupt();
                }
            }
        });
    }

    public static void main(String[] args) throws Exception {
        if (args.length < 2) {
            throw new IllegalArgumentException(
                    "This program takes two arguments: the path to an environment configuration file and" +
                            "the path to the file with records to send");
        }

        final Properties props = KafkaProducerApplication.loadProperties(args[0]);
        final String topic = props.getProperty("output.topic.name");
        final Producer<String, String> producer = new KafkaProducer<>(props);
        final KafkaProducerApplication producerApp = new KafkaProducerApplication(producer, topic);

        // Attach shutdown handler to catch Control-C.
        Runtime.getRuntime().addShutdownHook(new Thread(producerApp::shutdown));
        String filePath = args[1];
        try {
            List<String> linesToProduce = Files.readAllLines(Paths.get(filePath));
            List<Future<RecordMetadata>> metadata = linesToProduce.stream()
                    .filter(l -> !l.trim().isEmpty())
                    .map(producerApp::produce)
                    .collect(Collectors.toList());
            producerApp.printMetadata(metadata, filePath);

        } catch (IOException e) {
            System.err.println(String.format("Error reading file %s due to %s", filePath, e));
        }
        producerApp.shutdown();
    }
}

7
Create data to produce to Kafka

Create the following file input.txt in the base directory of the tutorial. The numbers before the - will be the key and the part after will be the value.

1-value
2-words
3-All Streams
4-Lead to
5-Kafka
6-Go to
7-Kafka Summit
8-How can
9-a 10 ounce
10-bird carry a
11-5lb coconut


8
Compile and run the KafkaProducer application

In your terminal, run:

./gradlew shadowJar

Now that you have an uberjar for the KafkaProducerApplication, you can launch it locally.

java -jar build/libs/kafka-producer-application-standalone-0.0.1.jar configuration/dev.properties input.txt

After you run the previous command, the application will process the file and you should something like this on the console:

Offsets and timestamps committed in batch from input.txt
Record written to offset 0 timestamp 1597352120029
Record written to offset 1 timestamp 1597352120037
Record written to offset 2 timestamp 1597352120037
Record written to offset 3 timestamp 1597352120037
Record written to offset 4 timestamp 1597352120037
Record written to offset 5 timestamp 1597352120037
Record written to offset 6 timestamp 1597352120037
Record written to offset 7 timestamp 1597352120037
Record written to offset 8 timestamp 1597352120037
Record written to offset 9 timestamp 1597352120037
Record written to offset 10 timestamp 1597352120038

Now you can experiment some by creating your own file in base directory and re-run the above command and substitute your file name for input.txt

Remember any data before the - is the key and data after is the value.

9
Confirm records sent by consuming from topic

Now run a console consumer that will read topics from the output topic to confirm your application published the expected records.

kafka-console-consumer --topic output-topic \
 --bootstrap-server broker:9092 \
 --from-beginning \
 --property print.key=true \
 --property key.separator=" : "

The output from the consumer can vary if you added any of your own records, but it should look something like this:

1 : value
2 : words
3 : All Streams
4 : Lead to
5 : Kafka
6 : Go to
7 : Kafka Summit
8 : How can
9 : a 10 ounce
10 : bird carry a
11 : 5lb coconut

Now close the consumer with a CTRL+C then the broker shell with a CTRL+D

Test it

1
Create a test configuration file

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

key.serializer=org.apache.kafka.common.serialization.StringSerializer
value.serializer=org.apache.kafka.common.serialization.StringSerializer
acks=all


#Properties below this line are specific to code in this application
input.topic.name=input-topic
output.topic.name=output-topic

2
Write a unit test

Create a directory for the tests to live in:

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

Next we will see why the thin wrapper class around the KafkaProducer makes testing easier. The KafkaProducerApplication accepts an instance of the Producer interface. The use of the interface allows us to inject any concrete type we want, including a Mock Producer for testing.

We use a MockProducer because you only want to test your own code. So you really only need to test that the producer recieves the expected records and in the expected format. Plus since there is no broker, the tests run very fast, which becomes an important factor as the number of tests increase.

There is only one method in KafkaProducerApplicationTest annotated with @Test, and that is testProduce(). Before you create the test, let’s go over a few of the key points of the test

final List<String> records = Arrays.asList("foo#bar", "bar#foo", "baz#bar", "great-weather");

records.forEach(producerApp::produce); (1)

final List<KeyValue<String, String>> expectedList = Arrays.asList(KeyValue.pair("foo", "bar"),
            KeyValue.pair("bar", "foo"),
            KeyValue.pair("baz", "bar"),
            KeyValue.pair("NO-KEY","great:weather")); (2)
final List<KeyValue<String, String>> actualList = mockProducer.history().stream().map(this::toKeyValue).collect(Collectors.toList()); (3)
1 Call the produce method
2 Build the expected list of records the producer should recieve
3 Use the MockProducer.history() method to get the records sent to the producer so the test can assert the expected records match the actual ones sent

Now create the following file at src/test/java/io/confluent/developer/KafkaProducerApplicationTest.java.

package io.confluent.developer;


import static org.hamcrest.CoreMatchers.equalTo;
import static org.junit.Assert.assertThat;

import java.io.IOException;
import java.util.Arrays;
import java.util.List;
import java.util.Properties;
import java.util.stream.Collectors;
import org.apache.kafka.clients.producer.MockProducer;
import org.apache.kafka.clients.producer.ProducerRecord;
import org.apache.kafka.streams.KeyValue;
import org.junit.Test;


public class KafkaProducerApplicationTest {

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

    @Test
    public void testProduce() throws IOException {
        final MockProducer<String, String> mockProducer = new MockProducer<>();
        final Properties props = KafkaProducerApplication.loadProperties(TEST_CONFIG_FILE);
        final String topic = props.getProperty("output.topic.name");
        final KafkaProducerApplication producerApp = new KafkaProducerApplication(mockProducer, topic);
        final List<String> records = Arrays.asList("foo-bar", "bar-foo", "baz-bar", "great:weather");

        records.forEach(producerApp::produce);

        final List<KeyValue<String, String>> expectedList = Arrays.asList(KeyValue.pair("foo", "bar"),
            KeyValue.pair("bar", "foo"),
            KeyValue.pair("baz", "bar"),
            KeyValue.pair("NO-KEY","great:weather"));

        final List<KeyValue<String, String>> actualList = mockProducer.history().stream().map(this::toKeyValue).collect(Collectors.toList());

        assertThat(actualList, equalTo(expectedList));
        producerApp.shutdown();
    }


    private KeyValue<String, String> toKeyValue(final ProducerRecord<String, String> producerRecord) {
        return KeyValue.pair(producerRecord.key(), producerRecord.value());
    }
}

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.

bootstrap.servers=<FILL ME IN>

key.serializer=org.apache.kafka.common.serialization.StringSerializer
value.serializer=org.apache.kafka.common.serialization.StringSerializer
acks=all


#Properties below this line are specific to code in this application
input.topic.name=<FILL ME IN>
output.topic.name=<FILL ME IN>

2
Build a Docker image

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

gradle jibDockerBuild --image=io.confluent.developer/kafka-producer-application-join: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/kafka-producer-application-join:0.0.1 config.properties

Deploy on Confluent Cloud

1
Run your app to Confluent Cloud

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

First, create your Kafka cluster in Confluent Cloud. Use the promo code CC100KTS to receive an additional $100 free usage (details).

Next, from the Confluent Cloud UI, click on Tools & client config 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.

Now you’re all set to your run application locally while your Kafka topics and stream processing is backed to your Confluent Cloud instance.