Learn the basics

ksqlDB Quickstart

The guide below demonstrates how to get a minimal environment up and running. Choose the distribution that's right for you.

When you're ready to learn more, take the next steps with use-case-driven tutorials.

Dive deep with use-case tutorials

Standalone

via Confluent Cloud

via Confluent Platform

The latest features, supported by the community.

Steps for:

1. Get standalone ksqlDB

Since ksqlDB runs natively on Apache Kafka®, you'll need to have a Kafka installation running that ksqlDB is configured to use. The docker-compose files to the right will run everything for you via Docker, including ksqlDB itself.

Select the docker-compose file that you'd like to use, depending on whether or not you're already running Kafka. Next, copy and paste it into a file named docker-compose.yml on your local filesystem.

---
version: '2'

services:
  zookeeper:
    image: confluentinc/cp-zookeeper:7.4.0
    hostname: zookeeper
    container_name: zookeeper
    ports:
      - "2181:2181"
    environment:
      ZOOKEEPER_CLIENT_PORT: 2181
      ZOOKEEPER_TICK_TIME: 2000

  broker:
    image: confluentinc/cp-kafka:7.4.0
    hostname: broker
    container_name: broker
    depends_on:
      - zookeeper
    ports:
      - "29092:29092"
    environment:
      KAFKA_BROKER_ID: 1
      KAFKA_ZOOKEEPER_CONNECT: 'zookeeper:2181'
      KAFKA_LISTENER_SECURITY_PROTOCOL_MAP: PLAINTEXT:PLAINTEXT,PLAINTEXT_HOST:PLAINTEXT
      KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://broker:9092,PLAINTEXT_HOST://localhost:29092
      KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR: 1
      KAFKA_GROUP_INITIAL_REBALANCE_DELAY_MS: 0
      KAFKA_TRANSACTION_STATE_LOG_MIN_ISR: 1
      KAFKA_TRANSACTION_STATE_LOG_REPLICATION_FACTOR: 1

  ksqldb-server:
    image: confluentinc/ksqldb-server:0.29.0
    hostname: ksqldb-server
    container_name: ksqldb-server
    depends_on:
      - broker
    ports:
      - "8088:8088"
    environment:
      KSQL_LISTENERS: http://0.0.0.0:8088
      KSQL_BOOTSTRAP_SERVERS: broker:9092
      KSQL_KSQL_LOGGING_PROCESSING_STREAM_AUTO_CREATE: "true"
      KSQL_KSQL_LOGGING_PROCESSING_TOPIC_AUTO_CREATE: "true"

  ksqldb-cli:
    image: confluentinc/ksqldb-cli:0.29.0
    container_name: ksqldb-cli
    depends_on:
      - broker
      - ksqldb-server
    entrypoint: /bin/sh
    tty: true

2. Start ksqlDB's server

From a directory containing the docker-compose.yml file created in the previous step, run this command in order to start all services in the correct order.

Once all services have successfully launched, you will have a ksqlDB server running and ready to use.

docker-compose up

3. Start ksqlDB's interactive CLI

ksqlDB runs as a server which clients connect to in order to issue queries.

Run this command to connect to the ksqlDB server and enter an interactive command-line interface (CLI) session.

docker exec -it ksqldb-cli ksql http://ksqldb-server:8088

4. Create a stream

The first thing we're going to do is create a stream. A stream essentially associates a schema with an underlying Kafka topic. Here's what each parameter in the CREATE STREAM statement does:

  • kafka_topic - Name of the Kafka topic underlying the stream. In this case it will be automatically created because it doesn't exist yet, but streams may also be created over topics that already exist.
  • value_format - Encoding of the messages stored in the Kafka topic. For JSON encoding, each row will be stored as a JSON object whose keys/values are column names/values. For example: {"profileId": "c2309eec", "latitude": 37.7877, "longitude": -122.4205}
  • partitions - Number of partitions to create for the locations topic. Note that this parameter is not needed for topics that already exist.

Check the documentation for more information about streams.

Copy and paste this statement into your interactive CLI session, and press enter to execute the statement.

CREATE STREAM riderLocations (profileId VARCHAR, latitude DOUBLE, longitude DOUBLE)
  WITH (kafka_topic='locations', value_format='json', partitions=1);

5. Create materialized views

We might also want to keep track of the latest location of the riders using a materialized view. For this we create a table currentLocation by issuing a SELECT statement over the previously created stream. Note that the table will be incrementally updated as new rider location data arrives. We use the LATEST_BY_OFFSET aggregate function to denote the fact that we are only interested in the latest location of a rider.

To make it more fun, let us also materialize a derived table (Table ridersNearMountainView) that captures how far the riders are from a given location or city.

Copy and paste those table statements into your interactive CLI session, and press enter to execute.

Check the documentation for more information about tables and materialized views.

CREATE TABLE currentLocation AS
  SELECT profileId,
         LATEST_BY_OFFSET(latitude) AS la,
         LATEST_BY_OFFSET(longitude) AS lo
  FROM riderlocations
  GROUP BY profileId
  EMIT CHANGES;

CREATE TABLE ridersNearMountainView AS
  SELECT ROUND(GEO_DISTANCE(la, lo, 37.4133, -122.1162), -1) AS distanceInMiles,
         COLLECT_LIST(profileId) AS riders,
         COUNT(*) AS count
  FROM currentLocation
  GROUP BY ROUND(GEO_DISTANCE(la, lo, 37.4133, -122.1162), -1);

6. Run a push query over the stream

Now, let us run a push query over the stream. Run the given query using your interactive CLI session.

This query will output all rows from the riderLocations stream whose coordinates are within 5 miles of Mountain View.

This is the first thing that may feel a bit unfamiliar to you, because the query will never return until it's terminated. It will perpetually push output rows to the client as events are written to the riderLocations stream.

Leave this query running in the CLI session for now. Next, we're going to write some data into the riderLocations stream so that the query begins producing output.

-- Mountain View lat, long: 37.4133, -122.1162
SELECT * FROM riderLocations
  WHERE GEO_DISTANCE(latitude, longitude, 37.4133, -122.1162) <= 5 EMIT CHANGES;

7. Start another CLI session

Since the CLI session from (5) is busy waiting for output from the push query, let's start another session that we can use to write some data into ksqlDB.

docker exec -it ksqldb-cli ksql http://ksqldb-server:8088

8. Populate the stream with events

Run each of the given INSERT statements within the new CLI session, and keep an eye on the CLI session from (5) as you do.

The push query will output matching rows in real time as soon as they're written to the riderLocations stream.

INSERT INTO riderLocations (profileId, latitude, longitude) VALUES ('c2309eec', 37.7877, -122.4205);
INSERT INTO riderLocations (profileId, latitude, longitude) VALUES ('18f4ea86', 37.3903, -122.0643);
INSERT INTO riderLocations (profileId, latitude, longitude) VALUES ('4ab5cbad', 37.3952, -122.0813);
INSERT INTO riderLocations (profileId, latitude, longitude) VALUES ('8b6eae59', 37.3944, -122.0813);
INSERT INTO riderLocations (profileId, latitude, longitude) VALUES ('4a7c7b41', 37.4049, -122.0822);
INSERT INTO riderLocations (profileId, latitude, longitude) VALUES ('4ddad000', 37.7857, -122.4011);

9. Run a Pull query against the materialized view

Finally, we run a pull query against the materialized view to retrieve all the riders that are currently within 10 miles from Mountain View.

In contrast to the previous push query which runs continuously, the pull query follows a traditional request-response model retrieving the latest result from the materialized view.

SELECT * from ridersNearMountainView WHERE distanceInMiles <= 10;
Learn more with end-to-end use cases
Tutorial
Materialized cache
Build and serve incrementally-updated stateful views.
Tutorial
Streaming ETL pipeline
Manipulate in-flight data to connect arbitrary sources and sinks.
Tutorial
Event-driven microservices
Trigger changes based on observed patterns of events in a stream.
Tutorials
Stream processing cookbook
Use these recipes to easily get started at any level.