KSQL: A SQL Streaming Engine for Apache Kafka

DZone 's Guide to

KSQL: A SQL Streaming Engine for Apache Kafka

In this article, we discuss how to use KSQL, a SQL streaming engine for Apache Kafka, to write more scalable, elastic, and fault-tolerant code.

· Big Data Zone ·
Free Resource

KSQL is a SQL streaming engine for Apache Kafka. It provides an easy-to-use, yet powerful interactive SQL interface for stream processing on Kafka, without the need to write code in a programming language like Java or Python. KSQL is scalable, elastic, and fault-tolerant. It supports a wide range of streaming operations, including data filtering, transformations, aggregations, joins, windowing, and sessionization.

What Is Streaming?

In stream processing, data is continuously processed, as new data become available for analyzing. Data is processed sequentially as an unbounded stream and may be pulled in by a “listening” analytics system as a record in key-value pairs.

Stream of data

Below are a few key features of KSQL processing:

  1. Per record stream processing with millisecond latency.
  2. Data filtering.
  3. Data transformation and conversions.
  4. Data enrichment with join.
  5. Data manipulation with scalar functions.
  6. Data analysis with stateful processing, aggregation, and windowing operation.

A client application can use the Kafka Streams API for Stream processing on Kafka topic data, and underneath the Kafka Streams API are Kafka producers and consumers.

KSQL queries do stream processing, which is an abstraction of Kafka stream API which can consume stream data that are structured eg Avro, JSON, DELIMITED

Now, let’s take a look at how we can query in KSQL:

  1. Start your confluent.
  2. Open KSQL CLI with the help of <confluent-home>/bin/ksql.
  3.  Create a STREAM pageviews_original from the Kafka topic pageviews, specifying the value_format of DELIMITED. Describe the new STREAM. Notice that KSQL created additional columns called ROWTIME, which corresponds to the Kafka message timestamp, and ROWKEY, which corresponds to the Kafka message key:

    • ksql> CREATE STREAM pageviews_original (viewtime bigint, userid varchar,
      pageid varchar) WITH (kafka_topic='pageviews',value_format='DELIMITED');

  1.  Create a users_original table from the Kafka topic users, specifying the value_format of JSON. Describe the new table:

    • ksql> CREATE TABLE users_original (registertime bigint, gender varchar,
      regionid varchar, userid varchar) WITH (kafka_topic='users',value_format=
  2.  Show stream and table using:


KStream vs KTable

A Stream is a sequence of structured data. Once an event is introduced into a stream, it is immutable, meaning that it can't be updated or deleted. A Table, on the other hand, represents the current situation based on the events coming from a stream and they are mutable.

  1.  Create a persistent query by using the CREATE STREAM keywords to precede the SELECT statement:

    • ksql> CREATE STREAM pageviews_female AS SELECT users_original.userid AS
      userid, pageid, regionid, gender FROM pageviews_original LEFT JOIN
      users_original ON pageviews_original.userid = users_original.userid
      WHERE gender = 'FEMALE';

  1.  Write KSQL to output topic:
    • CREATE STREAM pageviews_female_like_89 WITH (kafka_topic='pageviews_enrich
      ed_r8_r9', value_format='DELIMITED') AS SELECT * FROM pageviews_female WHERE
      regionid LIKE '%_8' OR regionid LIKE '%_9'
kafka ,scala ,ksql ,apache ,streaming ,tutorial

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}