How do you combine aggregate values, like `count`, from multiple streams into a single result?
You can run your application with Confluent Cloud.
In the Kafka Streams application, use a combination of cogroup and and aggregate methods as shown below.
final Aggregator<String, LoginEvent, LoginRollup> loginAggregator = new LoginAggregator();
final KGroupedStream<String, LoginEvent> appOneGrouped = appOneStream.groupByKey();
final KGroupedStream<String, LoginEvent> appTwoGrouped = appTwoStream.groupByKey();
final KGroupedStream<String, LoginEvent> appThreeGrouped = appThreeStream.groupByKey();
appOneGrouped.cogroup(loginAggregator)
.cogroup(appTwoGrouped, loginAggregator)
.cogroup(appThreeGrouped, loginAggregator)
.aggregate(() -> new LoginRollup(new HashMap<>()), Materialized.with(Serdes.String(), loginRollupSerde))
.toStream().to(totalResultOutputTopic, Produced.with(stringSerde, loginRollupSerde));
To get started, make a new directory anywhere you’d like for this project:
mkdir cogrouping-streams && cd cogrouping-streams
Next, create a directory for configuration data:
mkdir configuration
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.
After you log in to Confluent Cloud, 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.
From the Confluent Cloud Console, navigate to your Kafka cluster and then select Clients in the lefthand navigation. From the Clients view, create a new client and click Java to get the connection information customized to your cluster.
Create new credentials for your Kafka cluster and Schema Registry, writing in appropriate descriptions so that the keys are easy to find and delete later. The Confluent Cloud Console 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. |
This tutorial has some steps for Kafka topic management and producing and consuming events, for which you can use the Confluent Cloud Console or the Confluent CLI. Follow the instructions here to install the Confluent CLI, and then follow these steps connect the CLI to your Confluent Cloud cluster.
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 "idea"
id "eclipse"
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"
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.CogroupingStreams"
)
}
}
shadowJar {
archiveBaseName = "cogrouping-streams-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=cogrouping-streams
replication.factor=3
app-one.topic.name=app-one-topic
app-one.topic.partitions=6
app-one.topic.replication.factor=3
app-two.topic.name=app-two-topic
app-two.topic.partitions=6
app-two.topic.replication.factor=3
app-three.topic.name=app-three-topic
app-three.topic.partitions=6
app-three.topic.replication.factor=3
output.topic.name=output-topic
output.topic.partitions=6
output.topic.replication.factor=3
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
This tutorial uses 4 streams. The three input streams have a record type of LoginEvent used to represent a user logging into an application. The fourth stream is an output stream that writes a LoginRollup object out to a topic. In the next steps you’ll create the Avro schemas for these objects.
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/login-event.avsc to create the LoginEvent event:
{
"namespace": "io.confluent.developer.avro",
"type": "record",
"name": "LoginEvent",
"fields": [
{"name": "app_id", "type": "string"},
{"name": "user_id", "type": "string"},
{"name": "time", "type": "long"}
]
}Next create another schema file src/main/avro/login-rollup.avsc to create the LoginRollup for the cogrouping result:
{
"namespace": "io.confluent.developer.avro",
"type": "record",
"name": "LoginRollup",
"fields": [
{"name": "login_by_app_and_user", "type": {
"type": "map",
"values": {
"type": "map",
"values": {"type": "long"}
}
}
}
]
}
Because we will use an Avro schema in our Java code, we’ll need to compile it. The Gradle Avro plugin is a part of the build, so it will see your new Avro files, generate Java code for them, and compile those and all other Java sources. Run this command to get it all done:
./gradlew build
Create a directory for the Java files in this project:
mkdir -p src/main/java/io/confluent/developer
Before you create the Java class to run the Cogrouping example, let’s dive into the main point of this tutorial, how we use cogrouping:
final Aggregator<String, LoginEvent, LoginRollup> loginAggregator = new LoginAggregator();
final KGroupedStream<String, LoginEvent> appOneGrouped = appOneStream.groupByKey();
final KGroupedStream<String, LoginEvent> appTwoGrouped = appTwoStream.groupByKey();
final KGroupedStream<String, LoginEvent> appThreeGrouped = appThreeStream.groupByKey();
appOneGrouped.cogroup(loginAggregator)
.cogroup(appTwoGrouped, loginAggregator)
.cogroup(appThreeGrouped, loginAggregator)
.aggregate(() -> new LoginRollup(new HashMap<>()), Materialized.with(Serdes.String(), loginRollupSerde))
.toStream().to(totalResultOutputTopic, Produced.with(stringSerde, loginRollupSerde));
You’re using the cogrouping functionality here to get an overall grouping of logins per application. Kafka Streams creates this total grouping by using an Aggregator who knows how to extract records from each grouped stream. Your Aggregator instance here knows how to correctly combine each LoginEvent into the larger LoginRollup object. You’ll learn more about Aggregator in the next step.
Next, you have three input streams: appOneStream, appTwoStream, and appThreeStream. You need the intermediate object KGroupedStream, so you execute the groupByKey() method on each stream. For this tutorial, we have assumed the incoming records already have keys. In cases where records lack keys, you need to use a key-selecting method (selectKey(), map(), or groupBy()) to successfully group by key.
Now with your KGroupedStream objects, you start creating your larger aggregate by calling KGroupedStream.cogroup() on the first stream, using your Aggregator. This first step returns a CogroupedKStream instance. Then for each remaining KGroupedStream, you execute CogroupedKSteam.cogroup() using one of the KGroupedStream instances and the Aggregator you created previously. You repeat this sequence of calls for all of the KGroupedStream objects you want to combine into an overall aggregate.
For more background on cogrouping functionality in stream you can read the KIP-150 proposal.
Now go ahead and create the Java file at src/main/java/io/confluent/developer/CogroupingStreams.java.
package io.confluent.developer;
import io.confluent.common.utils.TestUtils;
import io.confluent.developer.avro.LoginEvent;
import io.confluent.developer.avro.LoginRollup;
import io.confluent.kafka.serializers.AbstractKafkaSchemaSerDeConfig;
import io.confluent.kafka.serializers.KafkaAvroDeserializer;
import io.confluent.kafka.serializers.KafkaAvroSerializer;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroSerde;
import java.io.FileInputStream;
import java.io.IOException;
import java.time.Duration;
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.concurrent.CountDownLatch;
import org.apache.avro.specific.SpecificRecord;
import org.apache.kafka.clients.admin.AdminClient;
import org.apache.kafka.clients.admin.NewTopic;
import org.apache.kafka.clients.producer.KafkaProducer;
import org.apache.kafka.clients.producer.ProducerConfig;
import org.apache.kafka.clients.producer.ProducerRecord;
import org.apache.kafka.clients.producer.Producer;
import org.apache.kafka.common.serialization.Serde;
import org.apache.kafka.common.serialization.Serdes;
import org.apache.kafka.common.serialization.StringSerializer;
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.Aggregator;
import org.apache.kafka.streams.kstream.Consumed;
import org.apache.kafka.streams.kstream.KGroupedStream;
import org.apache.kafka.streams.kstream.KStream;
import org.apache.kafka.streams.kstream.Materialized;
import org.apache.kafka.streams.kstream.Produced;
public class CogroupingStreams {
public Topology buildTopology(Properties allProps) {
final StreamsBuilder builder = new StreamsBuilder();
final String appOneInputTopic = allProps.getProperty("app-one.topic.name");
final String appTwoInputTopic = allProps.getProperty("app-two.topic.name");
final String appThreeInputTopic = allProps.getProperty("app-three.topic.name");
final String totalResultOutputTopic = allProps.getProperty("output.topic.name");
final Serde<String> stringSerde = Serdes.String();
final Serde<LoginEvent> loginEventSerde = getSpecificAvroSerde(allProps);
final Serde<LoginRollup> loginRollupSerde = getSpecificAvroSerde(allProps);
final KStream<String, LoginEvent> appOneStream = builder.stream(appOneInputTopic, Consumed.with(stringSerde, loginEventSerde));
final KStream<String, LoginEvent> appTwoStream = builder.stream(appTwoInputTopic, Consumed.with(stringSerde, loginEventSerde));
final KStream<String, LoginEvent> appThreeStream = builder.stream(appThreeInputTopic, Consumed.with(stringSerde, loginEventSerde));
final Aggregator<String, LoginEvent, LoginRollup> loginAggregator = new LoginAggregator();
final KGroupedStream<String, LoginEvent> appOneGrouped = appOneStream.groupByKey();
final KGroupedStream<String, LoginEvent> appTwoGrouped = appTwoStream.groupByKey();
final KGroupedStream<String, LoginEvent> appThreeGrouped = appThreeStream.groupByKey();
appOneGrouped.cogroup(loginAggregator)
.cogroup(appTwoGrouped, loginAggregator)
.cogroup(appThreeGrouped, loginAggregator)
.aggregate(() -> new LoginRollup(new HashMap<>()), Materialized.with(Serdes.String(), loginRollupSerde))
.toStream().to(totalResultOutputTopic, Produced.with(stringSerde, loginRollupSerde));
return builder.build();
}
static <T extends SpecificRecord> SpecificAvroSerde<T> getSpecificAvroSerde(final Properties allProps) {
final SpecificAvroSerde<T> specificAvroSerde = new SpecificAvroSerde<>();
specificAvroSerde.configure((Map)allProps, false);
return specificAvroSerde;
}
public void createTopics(final Properties allProps) {
try (final AdminClient client = AdminClient.create(allProps)) {
final List<NewTopic> topics = new ArrayList<>();
topics.add(new NewTopic(
allProps.getProperty("app-one.topic.name"),
Integer.parseInt(allProps.getProperty("app-one.topic.partitions")),
Short.parseShort(allProps.getProperty("app-one.topic.replication.factor"))));
topics.add(new NewTopic(
allProps.getProperty("app-two.topic.name"),
Integer.parseInt(allProps.getProperty("app-two.topic.partitions")),
Short.parseShort(allProps.getProperty("app-two.topic.replication.factor"))));
topics.add(new NewTopic(
allProps.getProperty("app-three.topic.name"),
Integer.parseInt(allProps.getProperty("app-three.topic.partitions")),
Short.parseShort(allProps.getProperty("app-three.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);
}
}
public Properties loadEnvProperties(String fileName) throws IOException {
final Properties allProps = new Properties();
final 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.");
}
final CogroupingStreams instance = new CogroupingStreams();
final Properties allProps = instance.loadEnvProperties(args[0]);
final Topology topology = instance.buildTopology(allProps);
instance.createTopics(allProps);
TutorialDataGenerator dataGenerator = new TutorialDataGenerator(allProps);
dataGenerator.generate();
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);
}
static class TutorialDataGenerator {
final Properties properties;
public TutorialDataGenerator(final Properties properties) {
this.properties = properties;
}
public void generate() {
properties.put(ProducerConfig.KEY_SERIALIZER_CLASS_CONFIG, StringSerializer.class);
properties.put(ProducerConfig.VALUE_SERIALIZER_CLASS_CONFIG, KafkaAvroSerializer.class);
try (Producer<String, LoginEvent> producer = new KafkaProducer<String, LoginEvent>(properties)) {
HashMap<String, List<LoginEvent>> entryData = new HashMap<>();
List<LoginEvent> messages1 = Arrays.asList(new LoginEvent("one", "Ted", 12456L),
new LoginEvent("one", "Ted", 12457L),
new LoginEvent("one", "Carol", 12458L),
new LoginEvent("one", "Carol", 12458L),
new LoginEvent("one", "Alice", 12458L),
new LoginEvent("one", "Carol", 12458L));
final String topic1 = properties.getProperty("app-one.topic.name");
entryData.put(topic1, messages1);
List<LoginEvent> messages2 = Arrays.asList(new LoginEvent("two", "Bob", 12456L),
new LoginEvent("two", "Carol", 12457L),
new LoginEvent("two", "Ted", 12458L),
new LoginEvent("two", "Carol", 12459L));
final String topic2 = properties.getProperty("app-two.topic.name");
entryData.put(topic2, messages2);
List<LoginEvent> messages3 = Arrays.asList(new LoginEvent("three", "Bob", 12456L),
new LoginEvent("three", "Alice", 12457L),
new LoginEvent("three", "Alice", 12458L),
new LoginEvent("three", "Carol", 12459L));
final String topic3 = properties.getProperty("app-three.topic.name");
entryData.put(topic3, messages3);
entryData.forEach((topic, list) ->
list.forEach(message ->
producer.send(new ProducerRecord<String, LoginEvent>(topic, message.getAppId(), message), (metadata, exception) -> {
if (exception != null) {
exception.printStackTrace(System.out);
} else {
System.out.printf("Produced record at offset %d to topic %s %n", metadata.offset(), metadata.topic());
}
})
)
);
}
}
}
}
The Aggregator you saw in the previous step constructs a map of maps: the count of logins per user, per application. Below is the core logic of the LoginAggregator.
Each call to Aggregator.apply retrieves the user login map for the given application id (or creates one if it doesn’t exist). From there, the Aggregator increments the login count for the given user.
final String userId = loginEvent.getUserId();
final Map<String, Map<String, Long>> allLogins = loginRollup.getLoginByAppAndUser();
final Map<String, Long> userLogins = allLogins.computeIfAbsent(appId, key -> new HashMap<>());
userLogins.compute(userId, (k, v) -> v == null ? 1L : v + 1L);
While you could add the Aggregator instance as an in-line lambda to the topology, creating a separate class allows you to test the aggregator in isolation.
Next, create the following file at src/main/java/io/confluent/developer/LoginAggregator.java.
package io.confluent.developer;
import io.confluent.developer.avro.LoginEvent;
import io.confluent.developer.avro.LoginRollup;
import java.util.HashMap;
import java.util.Map;
import org.apache.kafka.streams.kstream.Aggregator;
public class LoginAggregator implements Aggregator<String, LoginEvent, LoginRollup> {
@Override
public LoginRollup apply(final String appId,
final LoginEvent loginEvent,
final LoginRollup loginRollup) {
final String userId = loginEvent.getUserId();
final Map<String, Map<String, Long>> allLogins = loginRollup.getLoginByAppAndUser();
final Map<String, Long> userLogins = allLogins.computeIfAbsent(appId, key -> new HashMap<>());
userLogins.compute(userId, (k, v) -> v == null ? 1L : v + 1L);
return loginRollup;
}
}
In your terminal, run:
./gradlew shadowJar
Now that you have an uberjar for the Kafka Streams application, you can launch it locally. When you run the following, the prompt won’t return, because the application will run until you exit it. There is always another message to process, so streaming applications don’t exit until you force them.
The application for this tutorial includes a record generator to populate three topics with data.
java -jar build/libs/cogrouping-streams-standalone-0.0.1.jar configuration/dev.properties
Now that you have sent the login events, let’s run a consumer to read the cogrouped output from your streams application
confluent kafka topic consume output-topic -b --value-format avro
You should see something like this
{"login_by_app_and_user":{"one":{"Carol":3,"Alice":1,"Ted":2}}}
{"login_by_app_and_user":{"two":{"Carol":2,"Bob":1,"Ted":1}}}
{"login_by_app_and_user":{"three":{"Carol":1,"Bob":1,"Alice":2}}}
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 of the resources you created. Verify they are destroyed to avoid unexpected charges.
First, create a test file at configuration/test.properties:
application.id=cogrouping-streams
bootstrap.servers=localhost:29092
schema.registry.url=mock://cogrouping-streams-test
state.dir=cogrouping-test-state
app-one.topic.name=app-one-topic
app-one.topic.partitions=1
app-one.topic.replication.factor=1
app-two.topic.name=app-two-topic
app-two.topic.partitions=1
app-two.topic.replication.factor=1
app-three.topic.name=app-three-topic
app-three.topic.partitions=1
app-three.topic.replication.factor=1
output.topic.name=output-topic
output.topic.partitions=1
output.topic.replication.factor=1
Create a directory for the tests to live in:
mkdir -p src/test/java/io/confluent/developer
Create the following file at src/test/java/io/confluent/developer/LoginAggregatorTest.java.
This tests the Aggregator the Cogrouping operation uses. As I said previously, you can easily include an instance of the Aggregator in-line as a lambda in the original topology. But by having it as a stand alone class, you can easily test the Aggregator in a unit test.
package io.confluent.developer;
import org.junit.Test;
import java.util.HashMap;
import io.confluent.developer.avro.LoginEvent;
import io.confluent.developer.avro.LoginRollup;
import static org.hamcrest.MatcherAssert.assertThat;
import static org.hamcrest.Matchers.is;
public class LoginAggregatorTest {
@Test
public void shouldAggregateValues() {
final LoginAggregator loginAggregator = new LoginAggregator();
final LoginRollup loginRollup = new LoginRollup();
loginRollup.setLoginByAppAndUser(new HashMap<>());
final String appOne = "app-one";
final String appTwo = "app-two";
final String appThree = "app-three";
final String user1 = "user1";
final String user2 = "user2";
loginAggregator.apply(appOne, login(appOne, user1), loginRollup);
loginAggregator.apply(appTwo, login(appTwo, user1), loginRollup);
loginAggregator.apply(appThree, login(appThree, user1), loginRollup);
assertThat(loginRollup.getLoginByAppAndUser().get(appOne).get(user1), is(1L));
assertThat(loginRollup.getLoginByAppAndUser().get(appTwo).get(user1), is(1L));
assertThat(loginRollup.getLoginByAppAndUser().get(appThree).get(user1), is(1L));
loginAggregator.apply(appOne, login(appOne, user1), loginRollup);
loginAggregator.apply(appTwo, login(appTwo, user1), loginRollup);
assertThat(loginRollup.getLoginByAppAndUser().get(appOne).get(user1), is(2L));
assertThat(loginRollup.getLoginByAppAndUser().get(appTwo).get(user1), is(2L));
assertThat(loginRollup.getLoginByAppAndUser().get(appThree).get(user1), is(1L));
loginAggregator.apply(appOne, login(appOne, user2), loginRollup);
loginAggregator.apply(appTwo, login(appTwo, user2), loginRollup);
loginAggregator.apply(appThree, login(appThree, user2), loginRollup);
loginAggregator.apply(appOne, login(appOne, user1), loginRollup);
loginAggregator.apply(appTwo, login(appTwo, user1), loginRollup);
loginAggregator.apply(appThree, login(appThree, user1), loginRollup);
assertThat(loginRollup.getLoginByAppAndUser().get(appOne).get(user1), is(3L));
assertThat(loginRollup.getLoginByAppAndUser().get(appTwo).get(user1), is(3L));
assertThat(loginRollup.getLoginByAppAndUser().get(appThree).get(user1), is(2L));
assertThat(loginRollup.getLoginByAppAndUser().get(appOne).get(user2), is(1L));
assertThat(loginRollup.getLoginByAppAndUser().get(appTwo).get(user2), is(1L));
assertThat(loginRollup.getLoginByAppAndUser().get(appThree).get(user2), is(1L));
}
private LoginEvent login(String appId, String userId) {
return new LoginEvent(appId, userId, System.currentTimeMillis());
}
}Now create the following file at src/test/java/io/confluent/developer/CogroupingStreamsTest.java. Testing a Kafka streams application requires a bit of test harness code, but happily the org.apache.kafka.streams.TopologyTestDriver class makes this much more pleasant that it would otherwise be.
There is only one method in CogroupingStreamsTest annotated with @Test, and that is cogroupingTest(). This method actually runs our Streams topology using the TopologyTestDriver and some mocked data that is set up inside the test method.
package io.confluent.developer;
import static org.junit.Assert.assertEquals;
import io.confluent.developer.avro.LoginEvent;
import io.confluent.developer.avro.LoginRollup;
import io.confluent.kafka.streams.serdes.avro.SpecificAvroSerde;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Properties;
import java.util.TreeMap;
import org.apache.kafka.common.serialization.Deserializer;
import org.apache.kafka.common.serialization.Serde;
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.TestOutputTopic;
import org.apache.kafka.streams.Topology;
import org.apache.kafka.streams.TopologyTestDriver;
import org.junit.Test;
public class CogroupingStreamsTest {
private final static String TEST_CONFIG_FILE = "configuration/test.properties";
@Test
public void cogroupingTest() throws IOException {
final CogroupingStreams instance = new CogroupingStreams();
final Properties allProps = instance.loadEnvProperties(TEST_CONFIG_FILE);
final String appOneInputTopicName = allProps.getProperty("app-one.topic.name");
final String appTwoInputTopicName = allProps.getProperty("app-two.topic.name");
final String appThreeInputTopicName = allProps.getProperty("app-three.topic.name");
final String totalResultOutputTopicName = allProps.getProperty("output.topic.name");
final Topology topology = instance.buildTopology(allProps);
try (final TopologyTestDriver testDriver = new TopologyTestDriver(topology, allProps)) {
final Serde<String> stringAvroSerde = Serdes.String();
final SpecificAvroSerde<LoginEvent> loginEventSerde = CogroupingStreams.getSpecificAvroSerde(allProps);
final SpecificAvroSerde<LoginRollup> rollupSerde = CogroupingStreams.getSpecificAvroSerde(allProps);
final Serializer<String> keySerializer = stringAvroSerde.serializer();
final Deserializer<String> keyDeserializer = stringAvroSerde.deserializer();
final Serializer<LoginEvent> loginEventSerializer = loginEventSerde.serializer();
final TestInputTopic<String, LoginEvent> appOneInputTopic = testDriver.createInputTopic(appOneInputTopicName, keySerializer, loginEventSerializer);
final TestInputTopic<String, LoginEvent> appTwoInputTopic = testDriver.createInputTopic(appTwoInputTopicName, keySerializer, loginEventSerializer);
final TestInputTopic<String, LoginEvent> appThreeInputTopic = testDriver.createInputTopic(appThreeInputTopicName, keySerializer, loginEventSerializer);
final TestOutputTopic<String, LoginRollup> outputTopic = testDriver.createOutputTopic(totalResultOutputTopicName, keyDeserializer, rollupSerde.deserializer());
final List<LoginEvent> appOneEvents = new ArrayList<>();
appOneEvents.add(LoginEvent.newBuilder().setAppId("one").setUserId("foo").setTime(5L).build());
appOneEvents.add(LoginEvent.newBuilder().setAppId("one").setUserId("bar").setTime(6l).build());
appOneEvents.add(LoginEvent.newBuilder().setAppId("one").setUserId("bar").setTime(7L).build());
final List<LoginEvent> appTwoEvents = new ArrayList<>();
appTwoEvents.add(LoginEvent.newBuilder().setAppId("two").setUserId("foo").setTime(5L).build());
appTwoEvents.add(LoginEvent.newBuilder().setAppId("two").setUserId("foo").setTime(6l).build());
appTwoEvents.add(LoginEvent.newBuilder().setAppId("two").setUserId("bar").setTime(7L).build());
final List<LoginEvent> appThreeEvents = new ArrayList<>();
appThreeEvents.add(LoginEvent.newBuilder().setAppId("three").setUserId("foo").setTime(5L).build());
appThreeEvents.add(LoginEvent.newBuilder().setAppId("three").setUserId("foo").setTime(6l).build());
appThreeEvents.add(LoginEvent.newBuilder().setAppId("three").setUserId("bar").setTime(7L).build());
appThreeEvents.add(LoginEvent.newBuilder().setAppId("three").setUserId("bar").setTime(9L).build());
final Map<String, Map<String, Long>> expectedEventRollups = new TreeMap<>();
final Map<String, Long> expectedAppOneRollup = new HashMap<>();
final LoginRollup expectedLoginRollup = new LoginRollup(expectedEventRollups);
expectedAppOneRollup.put("foo", 1L);
expectedAppOneRollup.put("bar", 2L);
expectedEventRollups.put("one", expectedAppOneRollup);
final Map<String, Long> expectedAppTwoRollup = new HashMap<>();
expectedAppTwoRollup.put("foo", 2L);
expectedAppTwoRollup.put("bar", 1L);
expectedEventRollups.put("two", expectedAppTwoRollup);
final Map<String, Long> expectedAppThreeRollup = new HashMap<>();
expectedAppThreeRollup.put("foo", 2L);
expectedAppThreeRollup.put("bar", 2L);
expectedEventRollups.put("three", expectedAppThreeRollup);
sendEvents(appOneEvents, appOneInputTopic);
sendEvents(appTwoEvents, appTwoInputTopic);
sendEvents(appThreeEvents, appThreeInputTopic);
final List<LoginRollup> actualLoginEventResults = outputTopic.readValuesToList();
final Map<String, Map<String, Long>> actualRollupMap = new HashMap<>();
for (LoginRollup actualLoginEventResult : actualLoginEventResults) {
actualRollupMap.putAll(actualLoginEventResult.getLoginByAppAndUser());
}
final LoginRollup actualLoginRollup = new LoginRollup(actualRollupMap);
assertEquals(expectedLoginRollup, actualLoginRollup);
}
}
private void sendEvents(List<LoginEvent> events, TestInputTopic<String, LoginEvent> testInputTopic) {
for (LoginEvent event : events) {
testInputTopic.pipeInput(event.getAppId(), event);
}
}
}
Now run the test, which is as simple as:
./gradlew test