How can you count the number of events in a Kafka topic based on some criteria?
Suppose you have a topic with events that represent ticket sales for movies. In this tutorial, you'll see an example of 'groupby count' in Kafka Streams and ksqlDB. We'll write a program that calculates the total number of tickets sold per movie.
Use the count() method to apply the count aggregation.
To get started, make a new directory anywhere you’d like for this project:
Share your tutorial progressmkdir aggregate-count && cd aggregate-count
Next, create the following docker-compose.yml file to obtain Confluent Platform (for Kafka in the cloud, see Confluent Cloud):
---
version: '2'
services:
zookeeper:
image: confluentinc/cp-zookeeper:6.2.1
hostname: zookeeper
container_name: zookeeper
ports:
- "2181:2181"
environment:
ZOOKEEPER_CLIENT_PORT: 2181
ZOOKEEPER_TICK_TIME: 2000
broker:
image: confluentinc/cp-kafka:6.2.1
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
schema-registry:
image: confluentinc/cp-schema-registry:6.2.1
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
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:4.0.2"
}
}
plugins {
id "idea"
id "java"
id "com.google.cloud.tools.jib" version "3.1.1"
}
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:1.10.2"
implementation "org.slf4j:slf4j-simple:1.7.30"
implementation "org.apache.kafka:kafka-streams:2.8.1"
implementation "io.confluent:kafka-streams-avro-serde:6.2.1"
testImplementation "org.apache.kafka:kafka-streams-test-utils:2.8.1"
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.AggregatingCount"
)
}
}
shadowJar {
archiveBaseName = "kstreams-aggregating-count-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=aggregating-count-app
bootstrap.servers=127.0.0.1:29092
schema.registry.url=http://127.0.0.1:8081
input.topic.name=movie-ticket-sales
input.topic.partitions=1
input.topic.replication.factor=1
output.topic.name=movie-tickets-sold
output.topic.partitions=1
output.topic.replication.factor=1
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/ticket-sale.avsc for the ticket sale events:
{
"namespace": "io.confluent.developer.avro",
"type": "record",
"name": "TicketSale",
"fields": [
{"name": "title", "type": "string"},
{"name": "sale_ts", "type": "string"},
{"name": "ticket_total_value", "type": "int"}
]
}
Because this Avro schema is used in the Java code, it needs to compile it. Run the following:
./gradlew build
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/AggregatingCount.java. Let’s take a close look at the buildTopology() method, which uses the Kafka Streams DSL.
The first thing the method does is create an instance of StreamsBuilder, which is the helper object that lets us build our topology. With our builder in hand, we can apply the following methods:
Call the stream() method to create a KStream<String, TicketSale> object.
Since we can’t make any assumptions about the key of this stream, we have to repartition it explicitly. We use the map() method for that, creating a new KeyValue instance for each record, using the movie title as the new key.
Group the events by that new key by calling the groupByKey() method. This returns a KGroupedStream object.
Use the count() method to apply the count aggregation.
Use the toStream() method to produce the count results to the specified output topic.
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.KeyValue;
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.Grouped;
import org.apache.kafka.streams.kstream.Produced;
import java.io.FileInputStream;
import java.io.InputStream;
import java.io.IOException;
import java.time.Duration;
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 io.confluent.developer.avro.TicketSale;
import io.confluent.common.utils.TestUtils;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroSerde;
import static io.confluent.kafka.serializers.AbstractKafkaSchemaSerDeConfig.SCHEMA_REGISTRY_URL_CONFIG;
public class AggregatingCount {
private SpecificAvroSerde<TicketSale> ticketSaleSerde(final Properties allProps) {
final SpecificAvroSerde<TicketSale> serde = new SpecificAvroSerde<>();
Map<String, String> config = (Map)allProps;
serde.configure(config, false);
return serde;
}
public Topology buildTopology(Properties allProps,
final SpecificAvroSerde<TicketSale> ticketSaleSerde) {
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(), ticketSaleSerde))
// Set key to title and value to ticket value
.map((k, v) -> new KeyValue<>(v.getTitle(), v.getTicketTotalValue()))
// Group by title
.groupByKey(Grouped.with(Serdes.String(), Serdes.Integer()))
// Apply COUNT method
.count()
// Write to stream specified by outputTopic
.toStream().mapValues(v -> v.toString() + " tickets sold").to(outputTopic, Produced.with(Serdes.String(), Serdes.String()));
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 AggregatingCount().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());
allProps.put(StreamsConfig.CACHE_MAX_BYTES_BUFFERING_CONFIG, 0);
Topology topology = this.buildTopology(allProps, this.ticketSaleSerde(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.cleanUp();
streams.start();
latch.await();
} catch (Throwable e) {
System.exit(1);
}
System.exit(0);
}
}
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-aggregating-count-standalone-0.0.1.jar configuration/dev.properties
In a new terminal, run:
docker exec -i schema-registry /usr/bin/kafka-avro-console-producer --topic movie-ticket-sales --bootstrap-server broker:9092 --property value.schema="$(< src/main/avro/ticket-sale.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:
{"title":"Die Hard","sale_ts":"2019-07-18T10:00:00Z","ticket_total_value":12}
{"title":"Die Hard","sale_ts":"2019-07-18T10:01:00Z","ticket_total_value":12}
{"title":"The Godfather","sale_ts":"2019-07-18T10:01:31Z","ticket_total_value":12}
{"title":"Die Hard","sale_ts":"2019-07-18T10:01:36Z","ticket_total_value":24}
{"title":"The Godfather","sale_ts":"2019-07-18T10:02:00Z","ticket_total_value":18}
{"title":"The Big Lebowski","sale_ts":"2019-07-18T11:03:21Z","ticket_total_value":12}
{"title":"The Big Lebowski","sale_ts":"2019-07-18T11:03:50Z","ticket_total_value":12}
{"title":"The Godfather","sale_ts":"2019-07-18T11:40:00Z","ticket_total_value":36}
{"title":"The Godfather","sale_ts":"2019-07-18T11:40:09Z","ticket_total_value":18}
Leaving your original terminal running, open another to consume the events that have been filtered by your application:
docker exec -it broker /usr/bin/kafka-console-consumer --topic movie-tickets-sold --bootstrap-server broker:9092 --from-beginning --property print.key=true
After the consumer starts, you should see the following messages. Note that for every key (movie), a sequence of output records (count) is emitted. Each record represents an update to the count, which is sent on every movie event specifically because caching is disabled in the code with StreamsConfig.CACHE_MAX_BYTES_BUFFERING_CONFIG set to 0. Read more on Record caches in the DSL.
The consumer prompt will hang, waiting for more events to arrive. To continue studying the example, send more events through the input terminal prompt. Otherwise, you can Control-C to exit the process.
Die Hard 1
Die Hard 2
The Godfather 1
Die Hard 3
The Godfather 2
The Big Lebowski 1
The Big Lebowski 2
The Godfather 3
The Godfather 4
First, create a test file at configuration/test.properties:
application.id=aggregating-count-app
bootstrap.servers=127.0.0.1:29092
schema.registry.url=mock://SR_DUMMY_URL:8081
input.topic.name=movie-ticket-sales
input.topic.partitions=1
input.topic.replication.factor=1
output.topic.name=movie-tickets-sold
output.topic.partitions=1
output.topic.replication.factor=1
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/AggregatingCountTest.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.KeyValue;
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.ArrayList;
import java.util.Arrays;
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.TicketSale;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroSerde;
import static java.util.Arrays.asList;
public class AggregatingCountTest {
private final static String TEST_CONFIG_FILE = "configuration/test.properties";
private TopologyTestDriver testDriver;
private SpecificAvroSerde<TicketSale> makeSerializer(Properties allProps) {
SpecificAvroSerde<TicketSale> 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 shouldCountTicketSales() throws IOException {
AggregatingCount aggCount = new AggregatingCount();
Properties allProps = aggCount.loadEnvProperties(TEST_CONFIG_FILE);
String inputTopic = allProps.getProperty("input.topic.name");
String outputTopic = allProps.getProperty("output.topic.name");
final SpecificAvroSerde<TicketSale> ticketSaleSpecificAvroSerde = makeSerializer(allProps);
Topology topology = aggCount.buildTopology(allProps, ticketSaleSpecificAvroSerde);
testDriver = new TopologyTestDriver(topology, allProps);
Serializer<String> keySerializer = Serdes.String().serializer();
Deserializer<String> keyDeserializer = Serdes.String().deserializer();
final List<TicketSale>
input = asList(
new TicketSale("Die Hard", "2019-07-18T10:00:00Z", 12),
new TicketSale("Die Hard", "2019-07-18T10:01:00Z", 12),
new TicketSale("The Godfather", "2019-07-18T10:01:31Z", 12),
new TicketSale("Die Hard", "2019-07-18T10:01:36Z", 24),
new TicketSale("The Godfather", "2019-07-18T10:02:00Z", 18),
new TicketSale("The Big Lebowski", "2019-07-18T11:03:21Z", 12),
new TicketSale("The Big Lebowski", "2019-07-18T11:03:50Z", 12),
new TicketSale("The Godfather", "2019-07-18T11:40:00Z", 36),
new TicketSale("The Godfather", "2019-07-18T11:40:09Z", 18)
);
testDriver
.createInputTopic(inputTopic, keySerializer, ticketSaleSpecificAvroSerde.serializer())
.pipeValueList(input);
final String outputLabel = " tickets sold";
List<String> originalCounts = new ArrayList<String>(Arrays.asList("1", "2", "1", "3", "2", "1", "2", "3", "4"));
List<String> expectedOutput = originalCounts.stream().map(v -> v + outputLabel).collect(Collectors.toList());
final List<KeyValue<String, String>> keyValues =
testDriver
.createOutputTopic(outputTopic, keyDeserializer, Serdes.String().deserializer())
.readKeyValuesToList();
List<String> actualOutput;
actualOutput = keyValues
.stream()
.filter(record -> record.value != null)
.map(record -> record.value)
.collect(Collectors.toList());
// System.out.println(actualOutput);
Assert.assertEquals(expectedOutput, actualOutput);
}
@After
public void cleanup() {
testDriver.close();
}
}
Now run the test, which is as simple as:
./gradlew test
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=aggregating-count-app
bootstrap.servers=<< FILL ME IN >>
schema.registry.url=<< FILL ME IN >>
input.topic.name=movie-ticket-sales
input.topic.partitions=<< FILL ME IN >>
input.topic.replication.factor=<< FILL ME IN >>
output.topic.name=movie-tickets-sold
output.topic.partitions=<< FILL ME IN >>
output.topic.replication.factor=<< FILL ME IN >>
In your terminal, execute the following to invoke the Jib plugin to build an image:
gradle jibDockerBuild --image=io.confluent.developer/kstreams-aggregating-count:0.0.1
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-aggregating-count:0.0.1 config.properties
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 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.
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.