Change logs for Rudder 9.2
-
Security benchmarks
-
Derivation: create custom benchmarks from the standard one, with customized control points, etc.
-
Scheduling: On Linux, the benchmark schedule can be configured to be coarses than the agent’s own schedule.
-
PDF export of the results
-
The dashboard now has score breakdowns.
-
Change logs
-
A lot of items now have a "recent changes tab".
-
It is now possible to restore the previous state of specific items without rollbacking the whole configuration.
-
Authentication
-
TOTP-based MFA is now part of the local user management.
-
JWT tokens can now be encrypted.
-
More CSV exports: rules compliance, node compliance, score details in benchmarks.
-
The multi-tenants feature can now handle read-write users (with some limitations).
-
The webapp now runs as a dedicated system user. All interactions with the
configuration-repositorygit repository now need to be made asrudder-webapp. -
The inventory now runs as the last step of the agent run, allowing to trigger an inventory in the current run.
-
Agents are now available for the ARM64 architecture on Ubuntu 24.04 and 26.04 LTS.
Under the hood
-
Windows agent
-
The Windows agent now requires
netstandard2.0. -
The MSI installer hab been updated to Wix 6.
-
The agent is now started by a new
agentddaemon, replacing the scheduled task by default. -
The Rust modules now run using a new F#-based runner instead of CLIs.
💾 Installing, upgrading and testing
-
Install docs for Debian/Ubuntu, RHEL/CentOS and SLES