Earlier versions
In case you missed a release, review a list of features from previous versions.
What's new in 4.1.16
- RBAC (Role Based Access Control) Feature - Client can take leverage of role based access of the User Behavior Analytics (UBA) functionality. RBAC is divided into three roles namely, Admin, Tenant admin and Read-only.
- Upgraded the app base image to v3.0.11.
- Upgraded the java version of the machine learning app to v17.0.7.
- Fixed security vulnerabilities.
What's new in 4.1.15
- Updated the ncurses library to v6.1.10.
What's new in 4.1.14
- UBA now displays the tenant name in the UBA header area if multi-tenant configuration is detected.
- UBA now provides options to select and delete multiple users at once on the search results page.
- UBA now uses the correct IP address when the Encrypted App host is used in the NAT environment.
- Improved ML to use the correct IP address when the Encrypted App host is used in the NAT environment.
- The ML model now recovers if the build fails with error - "lost user Id lookup object".
- Added more files that are gathered by using the ML download logs function on the UBA help or support page.
- Fixed an issue in UBA triggering an unexpected error state when you are viewing the user details in the IBM® QRadar® Security Orchestration Automation and Response (SOAR) App.
- Fixed an issue where a query on a custom model in ML fails validation if that query contained the character sequence $'.
- Fixed an issue that crashed the UBAController process when a configuration is saved and if it is existing in the zookeeper.
What's new in 4.1.13
- Decay risk factor can now be set to zero, which disables reducing the overall risk score for users.
- Error messages that are related to installing or uninstalling the machine learning is now displayed for 30 seconds on the installer page.
- Fixed an issue that was preventing proper redirection to the Use-case Manager when you are viewing a tenant instance of UBA when logged in as an admin.
- Fixed an issue where the use of View User Details on the username fields
that contained
N/A
in QRadar that matched random users. - Fixed an issue that prevented the failed machine learning models to self-correct.
- Fixed security vulnerabilities.