クラウドで Kafka を活用 : Confluent なら10倍の効果が見込める理由 | 無料の電子ブックを入手

Announcing ksqlDB 0.22.0

Get started with Confluent, for free

Watch demo: Kafka streaming in 10 minutes

作成者 :

We’re pleased to announce ksqlDB 0.22.0! This release includes source streams and source tables as well as improved pull query (for key-range predicates) and push query performance. All of these features are now available in Confluent Cloud. We’re excited to share them with you, and we will cover the most notable changes in this blog post. Check out the changelog for a complete list of features and fixes.

Execute pull queries instantly with SOURCE tables

Prior to ksqlDB 0.22, pull queries were limited to tables that were created with a CREATE TABLE AS SELECT (CTAS) statement. As you may know, a CTAS statement executes a long-running Kafka Streams job in the background that maintains an internal state store (RocksDB) for table state changes. ksqlDB uses these internal state stores to serve pull queries. In contrast, a CREATE TABLE (CT) statement is a metadata operation that defines the table schema. However, data is only stored in the underlying topic but is not pulled into a state store; hence, this is why they weren’t supported in the past (see the figure below).

Pull query not supported

Because of the above requirement, if you want to look at topic content and query it using pull queries, you need to execute the following steps:

  1. Register your topic as a table
    CREATE TABLE agents (id STRING PRIMARY KEY, name STRING) 
    WITH (KAFKA_TOPIC=’agentsTopic’, FORMAT='JSON');
    
  2. Create a new table using the CTAS statement
    CREATE TABLE agents_aux AS SELECT id, name FROM agents EMIT CHANGES;
  3. Issue pull queries on the new table
    ksql> SELECT id, name FROM agents_aux WHERE id = ‘007’;
    +-----------------------+-----------------------------------+
    |ID                     |NAME                               |
    +-----------------------+-----------------------------------+
    |007                    |James Bond                         |
    Query terminated
    

These steps are a hassle if you have several tables that must be queryable with pull queries.

In ksqDB 0.22, we reduced the above steps by introducing the concept of source tables. A source table allows you to instantly run pull queries on tables without the need for an auxiliary CTAS statement, thus reducing the number of steps that you need to immediately start looking at the data of a topic.

Source tables are created by using the SOURCE keyword in the CREATE TABLE statement.

CREATE SOURCE TABLE agents (id STRING PRIMARY KEY, name STRING) 
WITH (KAFKA_TOPIC=’agentsTopic’, FORMAT='JSON');

ksql> SELECT id, name FROM agents WHERE id = ‘007’; +-------------------------------+-----------------------------------+ |ID |NAME | +-------------------------------+-----------------------------------+ |007 |James Bond | Query terminated

Source tables share the behavior of a CTAS statement where an internal state store (RocksDB) is created and maintained in order to provide pull queries capabilities.

Pull query supported

The SOURCE keyword also makes the table read-only, which is useful when you process data from upstream topics you don’t own and want to protect from modifications. You can also create read-only streams applying the SOURCE keyword when defining a stream: CREATE SOURCE STREAM <schema> WITH(...). Because source streams/tables are read-only and should be used for topics you don’t own, INSERT statements and DROP ... DELETE TOPIC statements are not supported. This is also true for tables created with CTAS statements that are marked as read-only.

Optimizations for range-scan expressions

Before this release of ksqlDB, pull queries with a WHERE clause using a range expression were implemented as full table scans, that is, all the records from a table were retrieved from the state store and then filtered within ksqlDB. With the 0.22 release, range queries on the primary key are now optimized to retrieve the exact range of records from the underlying state store, which is both faster–because no additional filtering is needed–and also more efficient in terms of I/O.

To illustrate the performance advantage of the range-scan optimization we ran a benchmark on a table with 10 million rows. The schema of the stream and the materialized view are shown below:

CREATE STREAM riderLocations2str (profileId STRING, latitude DOUBLE, longitude DOUBLE)
  WITH (kafka_topic='locations', value_format='json', partitions=1);

CREATE TABLE t2str AS
SELECT profileId, LATEST_BY_OFFSET(latitude) AS v1, LATEST_BY_OFFSET(longitude) AS v2
FROM riderlocations2str
GROUP BY profileId EMIT CHANGES

We generated 10 million records with monotonically increasing (profile) keys and issued the following query retrieving a subset of all the records smaller than a certain threshold:

SELECT profileId, v1, v2 FROM t2str WHERE profileId < '${rangekey}';

The figure below illustrates the advantage of the range-scan optimization compared to non-optimized implementation based on table scans:

Range-scan enabled vs disabled

As expected, the less selective the query is, the bigger the advantage of the range-scan optimization becomes.

Improved push queries

Push queries allow you to subscribe to a SQL query for changes. Before this release of ksqlDB, each query required launching a Kafka Streams application. In ksqlDB 0.22.0, a subset of push queries can now utilize a new architecture that allows them to run within a shared Kafka Streams application, and thus can be executed much more efficiently than before.

Get started with ksqlDB

Get started with ksqlDB today, via the standalone distribution or with Confluent, and join the community to ask questions and find new resources.

Get Started

  • Matthias is an Apache Kafka committer and PMC member, and works as a software engineer at Confluent. His focus is data stream processing in general, and thus he contributes to ksqlDB and Kafka Streams. Before joining Confluent, Matthias conducted research on distributed data stream processing systems at Humboldt-University of Berlin, were he received his Ph.D. Matthias is also a committer at Apache Flink and Apache Storm.

Get started with Confluent, for free

Watch demo: Kafka streaming in 10 minutes

このブログ記事は気に入りましたか?今すぐ共有