How to filter a stream of events

Question:

How do you filter messages in a Kafka topic to contain only those that you're interested in?

Example use case:

Consider a topic with events that represent book publications. In this tutorial, we'll write a program that creates a new topic which only contains the events for a particular author.

Share your tutorial progress

Hands-on code example:

Confluent Cloud ksqlDB Kafka Streams Basic Kafka

New to Confluent Cloud? Get started here.

Short Answer

Use the .filter() function as seen below. The filter method takes a boolean function of each record’s key and value. The function you give it determines whether to pass each event through to the next stage of the topology.

builder.stream(inputTopic, Consumed.with(Serdes.String(), publicationSerde))
        .filter((name, publication) -> "George R. R. Martin".equals(publication.getName()))
        .to(outputTopic, Produced.with(Serdes.String(), publicationSerde));

Run it

Provision your Kafka cluster

This tutorial requires access to an Apache Kafka cluster, and the quickest way to get started free is on Confluent Cloud, which provides Kafka as a fully managed service.

Take me to Confluent Cloud

After you log in to Confluent Cloud, 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).
Click on LEARN and follow the instructions to launch a Kafka cluster and to enable Schema Registry.

Initialize the project

Make a local directory anywhere you’d like for this project:

mkdir filter-events && cd filter-events

Next, create a directory for configuration data:

mkdir configuration

Write the cluster information into a local file

From the Confluent Cloud Console, navigate to your Kafka cluster. From the Clients view, get the connection information customized to your cluster (select Java).

Create new credentials for your Kafka cluster and Schema Registry, and then Confluent Cloud will show a configuration similar to below with your new credentials automatically populated (make sure show API keys is checked). Copy and paste it into a configuration/ccloud.properties file on your machine.

# Required connection configs for Kafka producer, consumer, and admin
bootstrap.servers={{ BOOTSTRAP_SERVERS }}
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={{ SR_URL }}
basic.auth.credentials.source=USER_INFO
basic.auth.user.info={{ SR_API_KEY }}:{{ SR_API_SECRET }}

Do not directly copy and paste the above configuration. You must copy it from the Confluent Cloud Console so that it includes your Confluent Cloud information and credentials.

Download and setup the Confluent CLI

This tutorial has some steps for Kafka topic management and/or reading from or writing to Kafka topics, for which you can use the Confluent Cloud Console or install the Confluent CLI. Instructions for installing Confluent CLI and configuring it to your Confluent Cloud environment is available from within the Confluent Cloud Console: navigate to your Kafka cluster, click on the CLI and tools link, and run through the steps in the Confluent CLI tab.

The CLI clients for Confluent Cloud (ccloud) and Confluent Platform (confluent v1.0) have been unified into a single client Confluent CLI confluent v2.0. This tutorial uses the unified Confluent CLI confluent v2.0 (ccloud client will continue to work until sunset on May 9, 2022, and you can read the migration instructions to the unified confluent CLI at https://docs.confluent.io/confluent-cli/current/migrate.html).

Configure the project

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

buildscript {
  repositories {
    mavenCentral()
  }
  dependencies {
    classpath "com.commercehub.gradle.plugin:gradle-avro-plugin:0.22.0"
    classpath "com.github.jengelman.gradle.plugins:shadow:6.1.0"
  }
}

plugins {
  id "java"
  id "com.google.cloud.tools.jib" version "3.2.0"
}

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

repositories {
  mavenCentral()


  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:1.11.0"
  implementation "org.slf4j:slf4j-simple:1.7.36"
  implementation "org.apache.kafka:kafka-streams:3.1.0"
  implementation "io.confluent:kafka-streams-avro-serde:7.1.0"
  testImplementation "org.apache.kafka:kafka-streams-test-utils:3.1.0"
  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.FilterEvents"
    )
  }
}

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

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

gradle wrapper

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

application.id=filtering-app
replication.factor=3

input.topic.name=publications
input.topic.partitions=6
input.topic.replication.factor=3

output.topic.name=filtered-publications
output.topic.partitions=6
output.topic.replication.factor=3

Update the properties file with Confluent Cloud information

Using the command below, append the contents of configuration/ccloud.properties (with your Confluent Cloud configuration) to configuration/dev.properties (with the application properties).

cat configuration/ccloud.properties >> configuration/dev.properties

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/publication.avsc for the publication events:

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

Because this Avro schema is used in the Java code, it needs 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/FilterEvents.java. Notice the buildTopology method, which uses the Kafka Streams DSL. The filter method takes a boolean function of each record’s key and value. The function you give it determines whether to pass each event through to the next stage of the topology. In this case, we’re only interested in books authored by George R. R. Martin.

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.Consumed;
import org.apache.kafka.streams.kstream.Produced;

import java.io.FileInputStream;
import java.io.InputStream;
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.concurrent.CountDownLatch;
import java.time.Duration;

import io.confluent.common.utils.TestUtils;
import io.confluent.developer.avro.Publication;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroSerde;

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

public class FilterEvents {

  private SpecificAvroSerde<Publication> publicationSerde(final Properties allProps) {
    final SpecificAvroSerde<Publication> serde = new SpecificAvroSerde<>();
    Map<String, String> config = (Map)allProps;
    serde.configure(config, false);
    return serde;
  }

  public Topology buildTopology(Properties allProps,
                                final SpecificAvroSerde<Publication> publicationSerde) {
    final StreamsBuilder builder = new StreamsBuilder();

    final String inputTopic = allProps.getProperty("input.topic.name");
    final String outputTopic = allProps.getProperty("output.topic.name");

    builder.stream(inputTopic, Consumed.with(Serdes.String(), publicationSerde))
        .filter((name, publication) -> "George R. R. Martin".equals(publication.getName()))
        .to(outputTopic, Produced.with(Serdes.String(), publicationSerde));

    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.topic.name"),
        Integer.parseInt(allProps.getProperty("input.topic.partitions")),
        Short.parseShort(allProps.getProperty("input.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 IOException {
    if (args.length < 1) {
      throw new IllegalArgumentException(
          "This program takes one argument: the path to an environment configuration file.");
    }

    new FilterEvents().runRecipe(args[0]);
  }

  private void runRecipe(final String configPath) throws IOException {
    final Properties allProps = new Properties();
    try (InputStream inputStream = new FileInputStream(configPath)) {
      allProps.load(inputStream);
    }
    allProps.put(StreamsConfig.APPLICATION_ID_CONFIG, allProps.getProperty("application.id"));
    allProps.put(StreamsConfig.STATE_DIR_CONFIG, TestUtils.tempDirectory().getPath());

    Topology topology = this.buildTopology(allProps, this.publicationSerde(allProps));
    this.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);

  }
}

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-filter-standalone-0.0.1.jar configuration/dev.properties

Produce events to the input topic

In a new terminal window, run the following command to start a Confluent CLI producer:

confluent kafka topic produce publications \
  --parse-key \
  --delimiter ":" \
  --value-format avro \
  --schema src/main/avro/publication.avsc

You will be prompted for the Confluent Cloud Schema Registry credentials as shown below, which you can find in the configuration/ccloud.properties configuration file. Look for the configuration parameter basic.auth.user.info, whereby the ":" is the delimiter between the key and secret.

Enter your Schema Registry API key:
Enter your Schema Registry API secret:

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:

"George R. R. Martin":{"name": "George R. R. Martin", "title": "A Song of Ice and Fire"}
"C.S. Lewis":{"name": "C.S. Lewis", "title": "The Silver Chair"}
"C.S. Lewis":{"name": "C.S. Lewis", "title": "Perelandra"}
"George R. R. Martin":{"name": "George R. R. Martin", "title": "Fire & Blood"}
"J. R. R. Tolkien":{"name": "J. R. R. Tolkien", "title": "The Hobbit"}
"J. R. R. Tolkien":{"name": "J. R. R. Tolkien", "title": "The Lord of the Rings"}
"George R. R. Martin":{"name": "George R. R. Martin", "title": "A Dream of Spring"}
"J. R. R. Tolkien":{"name": "J. R. R. Tolkien", "title": "The Fellowship of the Ring"}
"George R. R. Martin":{"name": "George R. R. Martin", "title": "The Ice Dragon"}

Enter Ctrl+C to exit.

Consume filtered events from the output topic

Run the following command to start a Confluent CLI consumer to view the distinct click events:

confluent kafka topic consume filtered-publications -b --value-format avro

Depending on the cadence and values you produce in the steps above, you should see messages similar to the following:

{"name":"George R. R. Martin","title":"A Song of Ice and Fire"}
{"name":"George R. R. Martin","title":"Fire & Blood"}
{"name":"George R. R. Martin","title":"A Dream of Spring"}
{"name":"George R. R. Martin","title":"The Ice Dragon"}

Enter Ctrl+C to exit.

Teardown Confluent Cloud resources

You may try another tutorial, but if you don’t plan on doing other tutorials, use the Confluent Cloud Console or CLI to destroy all the resources you created. Verify they are destroyed to avoid unexpected charges.

Test it

Create a test configuration file

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

application.id=filtering-app
bootstrap.servers=127.0.0.1:29092
schema.registry.url=mock://SR_CLOUD_DUMMY_URL:8081

input.topic.name=publications
input.topic.partitions=1
input.topic.replication.factor=1

output.topic.name=filtered-publications
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/FilterEventsTest.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.Topology;
import org.apache.kafka.streams.TopologyTestDriver;
import org.junit.After;
import org.junit.Assert;
import org.junit.Test;

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

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

import static java.util.Arrays.asList;

public class FilterEventsTest {

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

  private TopologyTestDriver testDriver;

  private SpecificAvroSerde<Publication> makeSerializer(Properties allProps) {

    SpecificAvroSerde<Publication> serde = new SpecificAvroSerde<>();

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

    return serde;
  }

  @Test
  public void shouldFilterGRRMartinsBooks() throws IOException {
    FilterEvents fe = new FilterEvents();
    Properties allProps = fe.loadEnvProperties(TEST_CONFIG_FILE);

    String inputTopic = allProps.getProperty("input.topic.name");
    String outputTopic = allProps.getProperty("output.topic.name");

    final SpecificAvroSerde<Publication> publicationSpecificAvroSerde = makeSerializer(allProps);

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

    Serializer<String> keySerializer = Serdes.String().serializer();
    Deserializer<String> keyDeserializer = Serdes.String().deserializer();

    // Fixture
    Publication iceAndFire = new Publication("George R. R. Martin", "A Song of Ice and Fire");
    Publication silverChair = new Publication("C.S. Lewis", "The Silver Chair");
    Publication perelandra = new Publication("C.S. Lewis", "Perelandra");
    Publication fireAndBlood = new Publication("George R. R. Martin", "Fire & Blood");
    Publication theHobbit = new Publication("J. R. R. Tolkien", "The Hobbit");
    Publication lotr = new Publication("J. R. R. Tolkien", "The Lord of the Rings");
    Publication dreamOfSpring = new Publication("George R. R. Martin", "A Dream of Spring");
    Publication fellowship = new Publication("J. R. R. Tolkien", "The Fellowship of the Ring");
    Publication iceDragon = new Publication("George R. R. Martin", "The Ice Dragon");
    // end Fixture

    final List<Publication>
        input = asList(iceAndFire, silverChair, perelandra, fireAndBlood, theHobbit, lotr, dreamOfSpring, fellowship,
                       iceDragon);

    final List<Publication> expectedOutput = asList(iceAndFire, fireAndBlood, dreamOfSpring, iceDragon);

    testDriver.createInputTopic(inputTopic, keySerializer, publicationSpecificAvroSerde.serializer())
        .pipeValueList(input);

    List<Publication> actualOutput =
        testDriver
            .createOutputTopic(outputTopic, keyDeserializer, publicationSpecificAvroSerde.deserializer())
            .readValuesToList()
            .stream()
            .filter(Objects::nonNull)
            .collect(Collectors.toList());

    Assert.assertEquals(expectedOutput, actualOutput);
  }

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

}

Invoke the tests

Now run the test, which is as simple as:

./gradlew test

Take it to production

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=filtering-app
bootstrap.servers=<< FILL ME IN >>
schema.registry.url=<< FILL ME IN >>

input.topic.name=publications
input.topic.partitions=<< FILL ME IN >>
input.topic.replication.factor=<< FILL ME IN >>

output.topic.name=filtered-publications
output.topic.partitions=<< FILL ME IN >>
output.topic.replication.factor=<< FILL ME IN >>

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/kstreams-filter:0.0.1

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-filter:0.0.1 config.properties