You are viewing the RapidMiner Radoop documentation for version 10.3 - Check here for latest version
What's New in RapidMiner Radoop 10.3.1?
Released: November 1, 2023
New Features
- Radoop Proxy connections now support additional trusted certificates from your company. These certificates will be utilized for all SSL communication with Hadoop services. This eliminates the need to reinstall these, often autogenerated, certificates on every end-user machine or AI Hub instance. Going forward, they can be shared via the Radoop Proxy connection, conveniently stored in an AI Hub project.
- Introducing multi-homed cluster support by incorporating DNS setup into the Radoop connection. The internal IP <-> DNS mappings can be easily shared with end-users through a Radoop Connection stored in the AI Hub project. This eliminates the manual editing of host files on end-user machines or AI Hub instance, making the setup process more seamless.
- Additional Radoop connection test have been added to verify the SOCKS5 connectivity of Radoop Proxy.
- Existing Radoop connection test has been extended to identify issues with Kerberos client configuration and to cover the pure Kerberos login process.
Bugfixes
- [Fixed] Improved the Radoop Connection Test operator, enabling it to run outside of Radoop Nest. This enhancement made connection problem troubleshooting more informative, especially regarding fundamental networking or Kerberos authentication issues.
- [Fixed] Addressed Kerberos token relogin failures for long-running processes and UI sessions. This was achieved by appropriately isolating Hadoop security contexts as well as Radoop Proxy sessions.
Compatibility
Radoop 10.3.1 requires RapidMiner Studio and RapidMiner AI Hub version 10.2.0 (or later) to function properly. Additionally, due to the Java 11 requirement on the cluster side, Radoop 10.3.1 requires Hive 3.x to run on the Java 11 JVM. Both a Java 11 JVM and Spark 3.x need to be available on all computational nodes. Model scoring related functionality is not expected to work with Hadoop clusters that do not support Hive on Java 11, such as EMR 5.x, EMR 6.x, HDInsight 4.x, and 5.x.