ksqlDB for Confluent Platform Java Client

ksqlDB ships with a lightweight Java client that enables sending requests easily to a ksqlDB server from within your Java application, as an alternative to using the REST API. The client supports pull and push queries; inserting new rows of data into existing ksqlDB streams; creation and management of new streams, tables, and persistent queries; and also admin operations such as listing streams, tables, and topics.

Tip

The code for the ksqlDB Java client is available in the ksql repo in the ksqldb-api-client directory.

The client sends requests to the HTTP2 server endpoints. Pull and push queries are served by the /query-stream endpoint, and inserts are served by the /inserts-stream endpoint. All other requests are served by the /ksql endpoint. The client is compatible only with ksqlDB deployments that are on version 0.10.0 or later.

Use the Java client to:

Get started below or skip to the end for full examples.

Get started

Start by creating a pom.xml for your Java application:

<?xml version="1.0" encoding="UTF-8"?>

<project xmlns="http://maven.apache.org/POM/4.0.0"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
    <modelVersion>4.0.0</modelVersion>

    <groupId>my.ksqldb.app</groupId>
    <artifactId>my-ksqldb-app</artifactId>
    <version>0.0.1</version>

    <properties>
        <!-- Keep versions as properties to allow easy modification -->
        <java.version>8</java.version>
        <ksqldb.version>0.29.0</ksqldb.version>
        <!-- Maven properties for compilation -->
        <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
        <project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
    </properties>

    <repositories>
        <repository>
            <id>ksqlDB</id>
            <name>ksqlDB</name>
            <url>https://ksqldb-maven.s3.amazonaws.com/maven/</url>
        </repository>
    </repositories>

    <pluginRepositories>
        <pluginRepository>
            <id>ksqlDB</id>
            <url>https://ksqldb-maven.s3.amazonaws.com/maven/</url>
        </pluginRepository>
    </pluginRepositories>

    <dependencies>
        <dependency>
            <groupId>io.confluent.ksql</groupId>
            <artifactId>ksqldb-api-client</artifactId>
            <version>${ksqldb.version}</version>
            <classifier>with-dependencies</classifier>
        </dependency>
    </dependencies>

    <build>
        <plugins>
            <plugin>
              <groupId>org.apache.maven.plugins</groupId>
              <artifactId>maven-compiler-plugin</artifactId>
              <version>3.8.1</version>
              <configuration>
                <source>${java.version}</source>
                <target>${java.version}</target>
                <compilerArgs>
                  <arg>-Xlint:all</arg>
                </compilerArgs>
              </configuration>
            </plugin>
        </plugins>
    </build>
</project>

Note

If you’re using ksqlDB for Confluent Platform (CP), use the Confluent Platform-specific modules from https://packages.confluent.io/maven/ by replacing repositories in the example POM above with a repository with this URL instead. Also update ksqldb.version to be a Confluent Platform version, such as 7.7.1, instead.

Note

The with-dependencies artifact was introduced in ksqlDB version 0.29 and Confluent Platform version 7.4.0. This jar includes all the necessary dependencies and relocates most of them in an attempt to avoid classpath issues. Using this jar provides the easiest way to get started. If you want more control over the dependencies on the classpath, you can depend directly on the client using this dependency block instead:

<dependency>
  <groupId>io.confluent.ksql</groupId>
  <artifactId>ksqldb-api-client</artifactId>
  <version>${ksqldb.version}</version>
</dependency>``

If you do this, you will need to add all of the transitive dependencies for ksqldb-api-client.

Create your example app at src/main/java/my/ksqldb/app/ExampleApp.java:

package my.ksqldb.app;

import io.confluent.ksql.api.client.Client;
import io.confluent.ksql.api.client.ClientOptions;

public class ExampleApp {

  public static String KSQLDB_SERVER_HOST = "localhost";
  public static int KSQLDB_SERVER_HOST_PORT = 8088;

  public static void main(String[] args) {
    ClientOptions options = ClientOptions.create()
        .setHost(KSQLDB_SERVER_HOST)
        .setPort(KSQLDB_SERVER_HOST_PORT);
    Client client = Client.create(options);

    // Send requests with the client by following the other examples

    // Terminate any open connections and close the client
    client.close();
  }
}

For additional client options, see the ClientOptions.java class.

You can use the ClientOptions class to connect your Java client to Confluent Cloud. For more information, see Connect to a Confluent Cloud ksqlDB cluster.

Run the following command to compile the Java code:

mvn clean package

Run the following command to confirm that your code compiled:

jar -tf target/my-ksqldb-app-0.0.1.jar

Your output should resemble:

META-INF/
META-INF/MANIFEST.MF
my/
my/ksqldb/
my/ksqldb/app/
my/ksqldb/app/ExampleApp.class
META-INF/maven/
META-INF/maven/my.ksqldb.app/
META-INF/maven/my.ksqldb.app/my-ksqldb-app/
META-INF/maven/my.ksqldb.app/my-ksqldb-app/pom.xml
META-INF/maven/my.ksqldb.app/my-ksqldb-app/pom.properties

Receive query results one row at a time (streamQuery())

The streamQuery() method enables client apps to receive query results one row at a time, either asynchronously via a Reactive Streams subscriber or synchronously in a polling fashion.

You can use this method to issue both push and pull queries, but the usage pattern is better for push queries. For pull queries, consider using the executeQuery() method instead.

Query properties can be passed as an optional second argument. For more information, see the streamQuery() method in the Client.java class.

By default, push queries return only newly arriving rows. To start from the beginning of the stream or table, set the auto.offset.reset property to earliest.

public interface Client {

  /**
   * Executes a query (push or pull) and returns the results one row at a time.
   *
   * <p>This method may be used to issue both push and pull queries, but the usage
   * pattern is better for push queries. For pull queries, consider using the
   * {@link Client#executeQuery(String)} method instead.
   *
   * <p>If a non-200 response is received from the server, the {@code CompletableFuture} will be
   * failed.
   *
   * @param sql statement of query to execute
   * @return a future that completes once the server response is received, and contains the query
   *         result if successful
   */
  CompletableFuture<StreamedQueryResult> streamQuery(String sql);

  ...

}

Asynchronous Usage

To consume records asynchronously, create a Reactive Streams subscriber to receive query result rows:

import io.confluent.ksql.api.client.Row;
import org.reactivestreams.Subscriber;
import org.reactivestreams.Subscription;

public class RowSubscriber implements Subscriber<Row> {

  private Subscription subscription;

  public RowSubscriber() {
  }

  @Override
  public synchronized void onSubscribe(Subscription subscription) {
    System.out.println("Subscriber is subscribed.");
    this.subscription = subscription;

    // Request the first row
    subscription.request(1);
  }

  @Override
  public synchronized void onNext(Row row) {
    System.out.println("Received a row!");
    System.out.println("Row: " + row.values());

    // Request the next row
    subscription.request(1);
  }

  @Override
  public synchronized void onError(Throwable t) {
    System.out.println("Received an error: " + t);
  }

  @Override
  public synchronized void onComplete() {
    System.out.println("Query has ended.");
  }
}

Use the Java client to send the query result to the server and stream results to the subscriber:

client.streamQuery("SELECT * FROM MY_STREAM EMIT CHANGES;")
    .thenAccept(streamedQueryResult -> {
      System.out.println("Query has started. Query ID: " + streamedQueryResult.queryID());

      RowSubscriber subscriber = new RowSubscriber();
      streamedQueryResult.subscribe(subscriber);
    }).exceptionally(e -> {
      System.out.println("Request failed: " + e);
      return null;
    });

Synchronous Usage

To consume records one-at-a-time in a synchronous fashion, use the poll() method on the query result object. If poll() is called with no arguments, it blocks until a new row becomes available or the query is terminated. You can also pass a Duration argument to poll(), which causes poll() to return null if no new rows are received by the time the duration has elapsed. For more information, see the StreamedQueryResult.java class.

StreamedQueryResult streamedQueryResult = client.streamQuery("SELECT * FROM MY_STREAM EMIT CHANGES;").get();

for (int i = 0; i < 10; i++) {
  // Block until a new row is available
  Row row = streamedQueryResult.poll();
  if (row != null) {
    System.out.println("Received a row!");
    System.out.println("Row: " + row.values());
  } else {
    System.out.println("Query has ended.");
  }
}

Receive query results in a single batch (executeQuery())

The executeQuery() method enables client apps to receive query results as a single batch that’s returned when the query completes.

This method is suitable for both pull queries and for terminating push queries, for example, queries that have a LIMIT clause. For non-terminating push queries, use the streamQuery() method instead.

Query properties can be passed as an optional second argument. For more information, see the executeQuery() method in the Client.java class.

By default, push queries return only newly arriving rows. To start from the beginning of the stream or table, set the auto.offset.reset property to earliest.

public interface Client {

  /**
   * Executes a query (push or pull) and returns all result rows in a single batch, once the query
   * has completed.
   *
   * <p>This method is suitable for both pull queries and for terminating push queries,
   * for example, queries that have a {@code LIMIT} clause. For non-terminating push queries,
   * use the {@link Client#streamQuery(String)} method instead.
   *
   * @param sql statement of query to execute
   * @return query result
   */
  BatchedQueryResult executeQuery(String sql);

  ...

}

Example Usage

String pullQuery = "SELECT * FROM MY_MATERIALIZED_TABLE WHERE KEY_FIELD='some_key';";
BatchedQueryResult batchedQueryResult = client.executeQuery(pullQuery);

// Wait for query result
List<Row> resultRows = batchedQueryResult.get();

System.out.println("Received results. Num rows: " + resultRows.size());
for (Row row : resultRows) {
  System.out.println("Row: " + row.values());
}

Terminate a push query (terminatePushQuery())

The terminatePushQuery() method enables client apps to terminate push queries.

public interface Client {

  /**
   * Terminates a push query with the specified query ID.
   *
   * <p>If a non-200 response is received from the server, the {@code CompletableFuture} will be
   * failed.
   *
   * @param queryId ID of the query to terminate
   * @return a future that completes once the server response is received
   */
  CompletableFuture<Void> terminatePushQuery(String queryId);

  ...

}

The query ID is obtained from the query result response object when the client issues push queries, by using either the streamQuery() or executeQuery() methods.

Example Usage

Here’s an example of terminating a push query issued by using the streamQuery() method:

String pushQuery = "SELECT * FROM MY_STREAM EMIT CHANGES;";
StreamedQueryResult streamedQueryResult = client.streamQuery(pushQuery).get();

String queryId = streamedQueryResult.queryID();
client.terminatePushQuery(queryId).get();

Here’s an analogous example for terminating a push query issued by using the executeQuery() method:

String pushQuery = "SELECT * FROM MY_STREAM EMIT CHANGES LIMIT 10;";
BatchedQueryResult batchedQueryResult = client.executeQuery(pushQuery);

String queryId = batchedQueryResult.queryID().get();
client.terminatePushQuery(queryId).get();

Insert a new row into a stream (insertInto())

Client apps can insert a new row of data into an existing ksqlDB stream by using the insertInto() method. To insert multiple rows in a streaming fashion, see streamInserts() instead.

public interface Client {

  /**
   * Inserts a row into a ksqlDB stream.
   *
   * <p>The {@code CompletableFuture} will be failed if a non-200 response is received from the
   * server, or if the server encounters an error while processing the insertion.
   *
   * @param streamName name of the target stream
   * @param row the row to insert. Keys are column names and values are column values.
   * @return a future that completes once the server response is received
   */
  CompletableFuture<Void> insertInto(String streamName, KsqlObject row);

  ...

}

Rows for insertion are represented as KsqlObject instances. A KsqlObject represents a map of strings (in this case, column names) to values (column values).

Example Usage

Here’s an example of using the client to insert a new row into an existing stream ORDERS with schema (ORDER_ID BIGINT, PRODUCT_ID VARCHAR, USER_ID VARCHAR).

KsqlObject row = new KsqlObject()
    .put("ORDER_ID", 12345678L)
    .put("PRODUCT_ID", "UAC-222-19234")
    .put("USER_ID", "User_321");

client.insertInto("ORDERS", row).get();

Insert new rows in a streaming fashion (streamInserts())

Starting with ksqlDB 0.11.0, the streamInserts() method enables client apps to insert new rows of data into an existing ksqlDB stream in a streaming fashion. This is in contrast to the insertInto() method which inserts a single row per request.

public interface Client {

  /**
   * Inserts rows into a ksqlDB stream. Rows to insert are supplied by a
   * {@code org.reactivestreams.Publisher} and server acknowledgments are exposed similarly.
   *
   * <p>The {@code CompletableFuture} will be failed if a non-200 response is received from the
   * server.
   *
   * <p>See {@link InsertsPublisher} for an example publisher that may be passed an argument to
   * this method.
   *
   * @param streamName name of the target stream
   * @param insertsPublisher the publisher to provide rows to insert
   * @return a future that completes once the initial server response is received, and contains a
   *         publisher that publishes server acknowledgments for inserted rows.
   */
  CompletableFuture<AcksPublisher>
      streamInserts(String streamName, Publisher<KsqlObject> insertsPublisher);

  ...

}

Rows for insertion are represented as KsqlObject instances. A KsqlObject represents a map of strings (in this case, column names) to values (column values).

The rows to be inserted are supplied via a Reactive Streams publisher. For convenience, the Java client for ksqlDB ships with a simple publisher implementation suitable for use with the streamInserts() method out of the box. This implementation is the InsertsPublisher in the example usage below.

As the specified rows are inserted by the ksqlDB server, the server responds with acknowledgments that may be consumed from the AcksPublisher returned by the streamInserts() method. The AcksPublisher is a Reactive Streams publisher.

Example Usage

Here’s an example of using the client to insert new rows into an existing stream ORDERS, in a streaming fashion. The ORDERS stream has schema (ORDER_ID BIGINT, PRODUCT_ID VARCHAR, USER_ID VARCHAR).

InsertsPublisher insertsPublisher = new InsertsPublisher();
AcksPublisher acksPublisher = client.streamInserts("ORDERS", insertsPublisher).get();

for (long i = 0; i < 10; i++) {
  KsqlObject row = new KsqlObject()
      .put("ORDER_ID", i)
      .put("PRODUCT_ID", "super_awesome_product")
      .put("USER_ID", "super_cool_user");
  insertsPublisher.accept(row);
}
insertsPublisher.complete();

To consume server acknowledgments for the stream of inserts, implement a Reactive Streams subscriber to receive the acknowledgments:

import io.confluent.ksql.api.client.InsertAck;
import org.reactivestreams.Subscriber;
import org.reactivestreams.Subscription;

public class AcksSubscriber implements Subscriber<InsertAck> {

  private Subscription subscription;

  public AcksSubscriber() {
  }

  @Override
  public synchronized void onSubscribe(Subscription subscription) {
    System.out.println("Subscriber is subscribed.");
    this.subscription = subscription;

    // Request the first ack
    subscription.request(1);
  }

  @Override
  public synchronized void onNext(InsertAck ack) {
    System.out.println("Received an ack for insert number: " + ack.seqNum());

    // Request the next ack
    subscription.request(1);
  }

  @Override
  public synchronized void onError(Throwable t) {
    System.out.println("Received an error: " + t);
  }

  @Override
  public synchronized void onComplete() {
    System.out.println("Inserts stream has been closed.");
  }
}

and subscribe to the AcksPublisher from above:

acksPublisher.subscribe(new AcksSubscriber());

Create and manage new streams, tables, and persistent queries (executeStatement())

Starting with ksqlDB 0.11.0, the executeStatement() method enables client apps to:

  • Create new ksqlDB streams and tables
  • Drop existing ksqlDB streams and tables
  • Create new persistent queries, i.e., CREATE ... AS SELECT and INSERT INTO ... AS SELECT statements
  • Pause, Resume, and Terminate persistent queries
public interface Client {

  /**
   * Sends a SQL request to the ksqlDB server. This method supports 'CREATE', 'CREATE ... AS
   * SELECT', 'DROP', 'TERMINATE', and 'INSERT INTO ... AS SELECT' statements.
   *
   * <p>Each request should contain exactly one statement. Requests that contain multiple statements
   * will be rejected by the client, in the form of failing the {@code CompletableFuture}, and the
   * request will not be sent to the server.
   *
   * <p>The {@code CompletableFuture} is completed once a response is received from the server.
   * Note that the actual execution of the submitted statement is asynchronous, so the statement
   * may not have been executed by the time the {@code CompletableFuture} is completed.
   *
   * <p>If a non-200 response is received from the server, the {@code CompletableFuture} will be
   * failed.
   *
   * @param sql the request to be executed
   * @return a future that completes once the server response is received, and contains the query ID
   *         for statements that start new persistent queries
   */
  CompletableFuture<ExecuteStatementResult> executeStatement(String sql);

  ...

}

To use this method, pass in the SQL for the command to be executed. Query properties can be passed as an optional second argument. For more information, see the For more information, see the executeStatement() in the Client.java class.

As explained in the Javadocs for the method above, the CompletableFuture returned by the executeStatement() method is completed as soon as the ksqlDB server has accepted the statement and a response is received by the client. In most situations, the ksqlDB server will have already executed the statement by this time, but this is not guaranteed.

For statements that create new persistent queries, the query ID may be retrieved from the returned ExecuteStatementResult, as long as the ksqlDB server version is at least 0.11.0, and the statement has executed by the time the server response was completed.

Example Usage

Create a new ksqlDB stream, assuming the topic orders exists:

String sql = "CREATE STREAM ORDERS (ORDER_ID BIGINT, PRODUCT_ID VARCHAR, USER_ID VARCHAR)"
               + "WITH (KAFKA_TOPIC='orders', VALUE_FORMAT='json');";
client.executeStatement(sql).get();

Drop an existing ksqlDB table, assuming the table USERS exists:

client.executeStatement("DROP TABLE USERS;").get();

Start a persistent query that reads from the earliest offset, assuming the stream ORDERS exists:

String sql = "CREATE TABLE ORDERS_BY_USER AS "
               + "SELECT USER_ID, COUNT(*) as COUNT "
               + "FROM ORDERS GROUP BY USER_ID EMIT CHANGES;";
Map<String, Object> properties = Collections.singletonMap("auto.offset.reset", "earliest");
ExecuteStatementResult result = client.executeStatement(sql, properties).get();
System.out.println("Query ID: " + result.queryId().orElse("<null>"));

Terminate a persistent query, assuming a query with ID CTAS_ORDERS_BY_USER_0 exists:

client.executeStatement("TERMINATE CTAS_ORDERS_BY_USER_0;").get();

List streams, tables, topics, and queries

Starting with ksqlDB 0.11.0, the Java client for ksqlDB supports the following admin operations:

  • Listing ksqlDB streams, by using the listStreams() method
  • Listing ksqlDB tables, by using the listTables() method
  • Listing Kafka topics available for use with ksqlDB, by using the listTopics() method
  • Listing running ksqlDB queries, with the listQueries() method

Example Usage

List ksqlDB streams:

List<StreamInfo> streams = client.listStreams().get();
for (StreamInfo stream : streams) {
  System.out.println(
     stream.getName()
     + " " + stream.getTopic()
     + " " + stream.getKeyFormat()
     + " " + stream.getValueFormat()
     + " " + stream.isWindowed()
  );
}

List ksqlDB tables:

List<TableInfo> tables = client.listTables().get();
for (TableInfo table : tables) {
  System.out.println(
       table.getName()
       + " " + table.getTopic()
       + " " + table.getKeyFormat()
       + " " + table.getValueFormat()
       + " " + table.isWindowed()
    );
}

List Kafka topics:

List<TopicInfo> topics = client.listTopics().get();
for (TopicInfo topic : topics) {
  System.out.println(
       topic.getName()
       + " " + topic.getPartitions()
       + " " + topic.getReplicasPerPartition()
  );
}

List running ksqlDB queries:

List<QueryInfo> queries = client.listQueries().get();
for (QueryInfo query : queries) {
  System.out.println(query.getQueryType() + " " + query.getId());
  if (query.getQueryType() == QueryType.PERSISTENT) {
    System.out.println(query.getSink().get() + " " + query.getSinkTopic().get());
  }
}

For more information, see the Client.java class.

Describe specific streams and tables

Starting with ksqlDB 0.12.0, the describeSource() method enables client apps to fetch metadata for existing ksqlDB streams and tables. The metadata returned from this method includes the stream or table’s underlying topic name, column names and associated types, serialization formats, queries that read and write from the stream or table, and more. For more information, see the Client.java class.

Example Usage

Fetch metadata for the stream or table with name my_source:

SourceDescription description = client.describeSource("my_source").get();
System.out.println("This source is a " + description.type());
System.out.println("This stream/table has " + description.fields().size() + " columns.");
System.out.println(description.writeQueries().size() + " queries write to this stream/table.");
System.out.println(description.readQueries().size() + " queries read from this stream/table.");

Get metadata about the ksqlDB cluster

Starting with ksqlDB 0.16.0, the serverInfo() method enables client apps to fetch metadata about the ksqlDB cluster. The metadata returned from this method includes the version of ksqlDB the server is running, the Kafka cluster id and the ksqlDB service id. For more information, see the Client.java class.

Example Usage

Fetch server metadata:

ServerInfo serverInfo = client.serverInfo().get();
System.out.println("The ksqlDB version running on this server is " + serverInfo.getServerVersion());
System.out.println("The Kafka cluster this server is using is " + serverInfo.getKafkaClusterId());
System.out.println("The id of this ksqlDB service is " + serverInfo.getKsqlServiceId());

Manage, list and describe connectors

Starting with ksqlDB 0.18.0, the Java client for ksqlDB supports the following connector operations:

  • Creating new connectors by using the createConnector() method
  • Dropping existing connectors by using the dropConnector() method
  • Listing connectors by using the listConnectors() method
  • Describing a specific connector by using the describeConnector() method

Example Usage

Create a new connector:

Map<String, String> connectorProperties = ImmutableMap.of(
  "connector.class", "io.confluent.connect.jdbc.JdbcSourceConnector",
  "connection.url", "jdbc:postgresql://localhost:5432/my.db",
  "mode", "bulk",
  "topic.prefix", "jdbc-",
  "table.whitelist", "users",
  "key", "username"
);
client.createConnector("jdbc-connector", true, connectorProperties, false).get();

Drop a connector:

client.dropConnector("jdbc-connector", true).get();

List connectors:

List<ConnectorInfo> connectors = client.listConnectors().get();
for (ConnectorInfo connector : connectors) {
  System.out.println(connector.name()
    + " " + connector.type()
    + " " + connector.className()
    + " " + connector.state()
    + "\n"
  );
}

Describe a connector:

ConnectorDescription description = client.describeConnector("jdbc-connector").get();
System.out.println(description.name()
  + " is a " + description.type() + " connector.\n"
  + " The connector's class is " + description.className() + ".\n"
  + " The connector is currently " + description.state() + ".\n"
  + " It reads/writes to " + description.sources().size() + " ksqlDB sources"
  + " and uses " + description.topics().size() + " topics."
);

Define variables for substitution

Starting with ksqlDB 0.18.0, users can define session variables by calling the define() method and reference them in other functions by wrapping the variable name in ${}. The undefine() method undefines a session variable, and getVariables() returns a map of the currently defined variables and their values.

Substitution is supported for the following functions:

  • streamQuery
  • executeQuery
  • executeStatement
  • describeSource
  • createConnector
  • dropConnector
  • describeConnector

Example Usage

Define a new variable:

client.define("topic", "stream-topic");

Use a variable in executeStatement:

client.executeStatement("CREATE STREAM S (NAME STRING, AGE INTEGER) WITH (kafka_topic='${topic}', value_format='json');");

Undefine a variable:

client.undefine("topic");

Get all variables:

Map<String, Object> variables = client.getVariables();

Execute Direct HTTP Requests

Sometimes, you need to execute requests directly against the ksqlDB server for reasons including, but not limited to, accessing features in ksqlDB REST API that are not available in the API client, or deserializing responses into different classes that are more native to your application.

For this purpose, the Client now adds an HttpRequest and HttpResponse interface that you can use for sending direct requests.

Example Usage

Call the /info endpoint in ksqlDB with:

HttpResponse response = client.buildRequest("GET", "/info")
    .send()
    .get();

// check status with
assert response.status() == 200;

// parse body (a byte[]) with:
parseIntoJson(response.body())

// or use the helper method to read body into a map:
Map<String, Map<String, Object>> info = response.bodyAsMap();

Add query properties variables:

HttpResponse response = client.buildRequest("POST", "/ksql")
    .payload("ksql", "CREATE STREAM FOO AS CONCAT(A, `wow;`) FROM `BAR`;")
    .propertiesKey("streamsProperties")
    .property("auto.offset.reset", "earliest")
    .send()
    .get();
assert response.status() == 200;

Or build the entire payload manually:

HttpResponse response = client.buildRequest("POST", "/ksql")
    .payload("ksql", "CREATE STREAM FOO AS CONCAT(A, `wow;`) FROM `BAR`;")
    .payload("streamsProperties", Collections.singletonMap("auto.offset.reset", "earliest"))
    .send()
    .get();
assert response.status() == 200;

The send() method adds authentication headers as specified in ClientOptions.

Assert the existence of a topic or schema

Starting with ksqlDB 0.27, users can use the assertSchema and assertTopic methods to assert the existence of resources. If the assertion fails, then the method will return a failed CompletableFuture.

Example Usage

Assert that a topic and schema exist before creating a stream:

client.assertTopic("foo", ImmutableMap.of("partitions", 1, "replicas", 3), true)
        .thenCompose((a) -> client.assertSchema("foo-key", 3, true))
        .thenAccept((a) -> client.executeStatement("CREATE STREAM..."))
        .exceptionally(...);

Assert that a topic doesn’t exist:

client.assertTopic("foo", false)
        .thenAccept(() -> client.executeStatement("CREATE STREAM..."))
        .exceptionally(...);

Tutorial examples

Event-driven microservice

In the ksqlDB tutorial on creating an event-driven microservice, the ksqlDB CLI is used to create a stream for transactions, seed some transaction events, and process transaction events into a table and verify output. Here’s the equivalent functionality using the Java client for ksqlDB.

Create the transactions stream:

String sql = "CREATE STREAM transactions ("
             + "     tx_id VARCHAR KEY,"
             + "    email_address VARCHAR,"
             + "     card_number VARCHAR,"
             + "     timestamp VARCHAR,"
             + "     amount DECIMAL(12, 2)"
             + ") WITH ("
             + "     kafka_topic = 'transactions',"
             + "     partitions = 8,"
             + "     value_format = 'avro',"
             + "     timestamp = 'timestamp',"
             + "     timestamp_format = 'yyyy-MM-dd''T''HH:mm:ss'"
             + ");";
Map<String, Object> properties = Collections.singletonMap("auto.offset.reset", "earliest");
client.executeStatement(sql, properties).get();

Seed some transaction events:

// Create the rows to insert
List<KsqlObject> insertRows = new ArrayList<>();
insertRows.add(new KsqlObject()
   .put("EMAIL_ADDRESS", "michael@example.com")
   .put("CARD_NUMBER", "358579699410099")
   .put("TX_ID", "f88c5ebb-699c-4a7b-b544-45b30681cc39")
   .put("TIMESTAMP", "2020-04-22T03:19:58")
   .put("AMOUNT", new BigDecimal("50.25")));
insertRows.add(new KsqlObject()
   .put("EMAIL_ADDRESS", "derek@example.com")
   .put("CARD_NUMBER", "352642227248344")
   .put("TX_ID", "0cf100ca-993c-427f-9ea5-e892ef350363")
   .put("TIMESTAMP", "2020-04-25T12:50:30")
   .put("AMOUNT", new BigDecimal("18.97")));
insertRows.add(new KsqlObject()
   .put("EMAIL_ADDRESS", "colin@example.com")
   .put("CARD_NUMBER", "373913272311617")
   .put("TX_ID", "de9831c0-7cf1-4ebf-881d-0415edec0d6b")
   .put("TIMESTAMP", "2020-04-19T09:45:15")
   .put("AMOUNT", new BigDecimal("12.50")));
insertRows.add(new KsqlObject()
   .put("EMAIL_ADDRESS", "michael@example.com")
   .put("CARD_NUMBER", "358579699410099")
   .put("TX_ID", "044530c0-b15d-4648-8f05-940acc321eb7")
   .put("TIMESTAMP", "2020-04-22T03:19:54")
   .put("AMOUNT", new BigDecimal("103.43")));
insertRows.add(new KsqlObject()
   .put("EMAIL_ADDRESS", "derek@example.com")
   .put("CARD_NUMBER", "352642227248344")
   .put("TX_ID", "5d916e65-1af3-4142-9fd3-302dd55c512f")
   .put("TIMESTAMP", "2020-04-25T12:50:25")
   .put("AMOUNT", new BigDecimal("3200.80")));
insertRows.add(new KsqlObject()
   .put("EMAIL_ADDRESS", "derek@example.com")
   .put("CARD_NUMBER", "352642227248344")
   .put("TX_ID", "d7d47fdb-75e9-46c0-93f6-d42ff1432eea")
   .put("TIMESTAMP", "2020-04-25T12:51:55")
   .put("AMOUNT", new BigDecimal("154.32")));
insertRows.add(new KsqlObject()
   .put("EMAIL_ADDRESS", "michael@example.com")
   .put("CARD_NUMBER", "358579699410099")
   .put("TX_ID", "c5719d20-8d4a-47d4-8cd3-52ed784c89dc")
   .put("TIMESTAMP", "2020-04-22T03:19:32")
   .put("AMOUNT", new BigDecimal("78.73")));
insertRows.add(new KsqlObject()
   .put("EMAIL_ADDRESS", "colin@example.com")
   .put("CARD_NUMBER", "373913272311617")
   .put("TX_ID", "2360d53e-3fad-4e9a-b306-b166b7ca4f64")
   .put("TIMESTAMP", "2020-04-19T09:45:35")
   .put("AMOUNT", new BigDecimal("234.65")));
insertRows.add(new KsqlObject()
   .put("EMAIL_ADDRESS", "colin@example.com")
   .put("CARD_NUMBER", "373913272311617")
   .put("TX_ID", "de9831c0-7cf1-4ebf-881d-0415edec0d6b")
   .put("TIMESTAMP", "2020-04-19T09:44:03")
   .put("AMOUNT", new BigDecimal("150.00")));

// Insert the rows
List<CompletableFuture<Void>> insertFutures = new ArrayList<>();
for (KsqlObject row : insertRows) {
  insertFutures.add(client.insertInto("TRANSACTIONS", row));
}

// Wait for the inserts to complete
CompletableFuture<Void> allInsertsFuture =
    CompletableFuture.allOf(insertFutures.toArray(new CompletableFuture<?>[0]));
allInsertsFuture.thenRun(() -> System.out.println("Seeded transaction events."));

Create the anomalies tables:

String sql = "CREATE TABLE possible_anomalies WITH ("
             + "    kafka_topic = 'possible_anomalies',"
             + "    VALUE_AVRO_SCHEMA_FULL_NAME = 'io.ksqldb.tutorial.PossibleAnomaly'"
             + ")   AS"
             + "    SELECT card_number AS `card_number_key`,"
             + "           as_value(card_number) AS `card_number`,"
             + "           latest_by_offset(email_address) AS `email_address`,"
             + "           count(*) AS `n_attempts`,"
             + "           sum(amount) AS `total_amount`,"
             + "           collect_list(tx_id) AS `tx_ids`,"
             + "           WINDOWSTART as `start_boundary`,"
             + "           WINDOWEND as `end_boundary`"
             + "    FROM transactions"
             + "    WINDOW TUMBLING (SIZE 30 SECONDS, RETENTION 1000 DAYS)"
             + "    GROUP BY card_number"
             + "    HAVING count(*) >= 3"
             + "    EMIT CHANGES;";
Map<String, Object> properties = Collections.singletonMap("auto.offset.reset", "earliest");
client.executeStatement(sql, properties).get();

Check contents of the anomalies table with a push query:

String query = "SELECT * FROM possible_anomalies EMIT CHANGES;";
Map<String, Object> properties = Collections.singletonMap("auto.offset.reset", "earliest");
client.streamQuery(query, properties)
    .thenAccept(streamedQueryResult -> {
      System.out.println("Result column names: " + streamedQueryResult.columnNames());

      RowSubscriber subscriber = new RowSubscriber();
      streamedQueryResult.subscribe(subscriber);
    }).exceptionally(e -> {
      System.out.println("Push query request failed: " + e);
      return null;
    });

In the example above, RowSubscriber is the example subscriber implementation introduced in the section on the streamQuery() method above. The RowSubscriber implementation can be adapted to adjust how the received rows are printed, or to pass them to a downstream application.

Pull queries against a materialized view

As a second example, in the ksqlDB tutorial on building a materialized view/cache, the ksqlDB CLI is used to issue pull queries against materialized views containing information about customer calls to a call center. Here’s a similar set of queries using the Java client for ksqlDB:

String sql1 = "SELECT name, total_calls, minutes_engaged FROM lifetime_view WHERE name = 'derek';";
String sql2 = "SELECT name, total_calls, minutes_engaged FROM lifetime_view WHERE name = 'michael';";

// Execute two pull queries and compare the results
client.executeQuery(sql1).thenCombine(
    client.executeQuery(sql2),
    (queryResult1, queryResult2) -> {
      // One row is returned from each query, as long as the queried keys exist
      Row result1 = queryResult1.get(0);
      Row result2 = queryResult2.get(0);
      if (result1.getLong("TOTAL_CALLS") > result2.getLong("TOTAL_CALLS")) {
        System.out.println(result1.getString("NAME") + " made more calls.");
      } else {
        System.out.println(result2.getString("NAME") + " made more calls.");
      }
      return null;
    });

Connect to a Confluent Cloud ksqlDB cluster

Use the following code snippet to connect your Java client to a hosted ksqlDB cluster in Confluent Cloud.

ClientOptions options = ClientOptions.create()
 .setBasicAuthCredentials("<ksqlDB-API-key>", "<ksqlDB-API-secret>")
 .setHost("<ksqlDB-endpoint>")
 .setPort(443)
 .setUseTls(true)
 .setUseAlpn(true);

Get the API key and endpoint URL from your Confluent Cloud cluster.

Related content