Cloudera Docs

What's New in Streaming Analytics

Learn about the new Streaming Analytics features in Cloudera DataFlow for Data Hub 7.2.12.

The following new features are introduced in Streaming Analytics CDF for Data Hub 7.2.12:

Apache Flink upgrade
Apache Flink 1.13 is supported in Streaming Analytics 7.2.12 cluster templates.
Default integration for SSB user in Ranger
The SQL Stream Builder (SSB) service user is by default added for Apache Kafka and YARN in Apache Ranger. When preparing your environment, you do not have to add the SSB user to the Kafka and YARN policies.
SQL Session Cluster support
Session mode is supported for SQL Stream Builder which means that the Flink jobs are running in a session cluster by default. You are also able to set the properties for the session mode directly from the SQL window on the Streaming SQL Console.
For more information, see the Managing session for SQL jobs section.
Connector support
The following new connectors are added as Templates to SQL Stream Builder:
  • Change Data Capture (CDC) connectors: Oracle, MySQL, PostgreSQL
  • Blackhole connector
  • Faker connector
You are also able to use more types of Kafka connectors as there are individual templates for Upsert Kafka, Kafka in CDP and Kafka outside of CDP.
For more information, see the Supported tables in SQL Stream Builder section.
Streaming SQL Console updates
Expanded Flink SQL support
More SQL statements are available that can be used directly from the SQL window on the Streaming SQL Console.
INSERT INTO support
Instead of using a Sink Table on the Streaming SQL Console, you are able to use the INSERT INTO statement directly in the SQL window to output your query results.
Tables > Flink DDL option is no longer supported
Instead of using the Flink DDL wizard, you are able to use Templates under the SQL window to select and import the CREATE TABLE templates.
For more information, see the Concept of tables in SQL Stream Builder section.