Release 1.17.0

Highlights

End-user

• Flink Catalogs integration. Thanks to Catalogs, Nu can be used to act on data stored in Data Lakehouses (check our tutorial on Apache Iceberg integration).
• Improvements in batch mode, including joins, testing (dry runs), and general usability.
• Live counts. You can now watch counts of events as they flow through a scenario, with user defined update interval.
• Autocompletion now works better in invalid expressions. When working on an expression, you don't have to keep it valid all the time for autocompletion to work.
• SpEL collections usability improvements. Arrays and Lists are now unified (no need for conversions) and we also added more collection manipulation functions.

Administrator

• Performance optimizations around Avro processing that reduce CPU utilization significantly.
• Components configuration reloading without Designer restart. You can now reload the Components configuration on demand, using an endpoint, without restarting the whole application.

All changes

For a detailed list of changes see Changelog and Migration guide.

Release 1.16.3

All changes

For a detailed list of changes see Changelog

Release 1.16.2

All changes

For a detailed list of changes see Changelog

Release 1.16.1

All changes

For a detailed list of changes see Changelog

Release 1.16.0

Highlights

Administrator

• Added Exactly-Once semantics for event processing in streaming mode.
• Added impersonate mechanism. When an external system calls the Nussknacker API with a technical user, the impersonated user can be passed and the performed action will be carried on behalf of the impersonated user.
• Nussknacker Quickstart has been refurbished and is now based on installation example. Using Nussknacker Quickstart is now even clearer and easier.

All changes

For a detailed list of changes see Changelog and Migration guide.

Release 1.15.3

All changes

For a detailed list of changes see Changelog

Release 1.15.1

All changes

For a detailed list of changes see Changelog

Release 1.15.0

Highlights

End-user

• Experimental support for the batch processing mode, based on Flink’s Table API and Catalogs. It allows processing data at rest and opens up new features such as integration with Lakehouses
• New color theme, more consistent, readable, and more violet!
• Added node search in the scenario editor - you can now easily find specific nodes based on their properties

Administrator

• Support for Flink 1.18
• Added support for scenario migration between different Nussknacker versions. From now on, it will be possible to migrate scenarios using the migration API
• New one-command Docker installation script for setting up an example Nussknacker environment on-premise. This example can be a good reference for your own on-premises production setup, as well as for exploration and testing

All changes

For a detailed list of changes see Changelog and Migration guide.

Release 1.14.0

Highlights

End-user

• New Decision table component, that allows expressing multi-condition decisions in a straightforward way
• Scenario processing modes - streaming, request-response, or batch (in the next release!) - made a first-class citizen in Designer, to help users choose how to process the data
• Multiple scenario and fragment validation improvements

Administrator

• Added support for different topic naming conventions in Azure Schema Registries
• Bumped Flink to 1.17

Component developer

• Changes and improvements in Component API for custom component development

All changes

For a detailed list of changes see Changelog and Migration guide.

New Contributors

• @lukasz-bigorajski made their first contribution in #5334
• @Elmacioro made their first contribution in #5327

1.13.2

All changes

For a detailed list of changes see Changelog.