What's New in RapidMiner AI Hub 10.0?

Released: 3rd November 2022

This is a major version upgrade. Please ensure to read the Upgrade from AI Hub 9 migration instructions!

The server service has been modernized and overhauled massively and now comes with the latest tech stack, minimizing security risks and boosting the performance.

  • We have upgraded our platform to use Java 11 and latest Spring Boot working under the hood.
  • The Active MQ broker is replaced with Active MQ Artemis.
  • The frontend has been migrated to a more user-friendly theme using React and as single page application it also responds much faster.

Beyond the technical improvements we have also enhanced useability:

  • Keycloak is now the main authentication source which allows easier integration with external services like LDAP
  • We are providing now a back reference of which schedule created which job
  • We have implemented optimizations in regard to indexes in the used database tables
  • We have improved schedules functionality
    • "Pause all" and "Resume all" functionality is now available for Schedules
    • Toggling scheduler into standby is available
  • Configurable clean up mechanism for RTS execution
  • Configuration changes via environment variables, so messing with property files is now mostly a thing of the past

With the modernization we have made some strategic adjustments to the server's architecture in general.

  • the service is now only available as part of the overall containerized AI Hub. This means the server service authentication system has been removed.
  • the only supported central service database is Postgres which is built into the overall platform. NOTE: this does not affect any database connectivity for data retrieval and storage
  • the web service execution has to be done via Real-Time Scoring Agent deployments using projects
  • Support for legacy repository has been removed and Projects are now the storage place for assets in AI Hub 10
  • web apps are not available any more. The required charting and monitoring can be setup via the built-in Grafana service.