Behavioral Changes in Streaming Analytics

Review the list of Streaming Analytics behavioral changes in Cloudera DataFlow for Data Hub 7.2.18.

SQL Stream Builder

Summary:
The user-defined function (UDF) tester is not available on Streaming SQL Console
Previous behavior:
New UDFs could be tested on Streaming SQL Console before executing the SQL job.
New behavior:
The UDF test runner is turned off, the added UDFs can be tested with executing the SQL job.
Summary:
Kafka sink JSON serializer is refactored to the Flink implementation
Previous behavior:
An SSB specific implementation was used for the JSON serializer.
New behavior:
The RowJsonSerializer is added to be used for the Kafka sink JSON serializer. The v2 API version is also added for the JSONSampler.
Summary:
Remove user name length limitation
Previous behavior:
Username limitation was set to a minimum four characters.
New behavior:
Limitation is removed for username length.
Summary:
Make API v2 the default selection in the SSB Swagger UI
Previous behavior:
Version could be selected for the REST API when using Swagger UI.
New behavior:
V2 is the default version when using REST API with Swagger UI.
Summary:
ssb-hive-catalog can be configured when creating Iceberg catalog
Previous behavior:
ssb-hive-catalog could only be configured for creating Iceberg tables
New behavior:
ssb-hive-catalog can be used both for creating Iceberg catalogs and tables for seamless Hive integration.

Flink

Summary:
Remove global JobResultStore storage path
Previous behavior:
The JobResultStore was configured as a global HDFS path.
New behavior:
JobResultStore is automatically provided by YARN.