Skip to main content

Managed Addons

Managed Addons help manage additional components required for your LLMOS cluster. By default, LLMOS comes with a set of system addons that are enabled out of the box. These system addons include:

  • Nvidia GPU Operator: Enables support for Nvidia GPUs in your LLMOS cluster using the Nvidia GPU Operator.
  • KubeRay Operator: Facilitates the management of Ray clusters and other machine learning resources in your LLMOS cluster through the KubeRay Operator.
  • LLMOS Redis: Provides a LLMOS-managed Redis service for your cluster.
  • Rook Ceph Operator: Adds storage support to your cluster using the Rook Ceph operator.
  • System Update Controller: Manages LLMOS node upgrades with the System Upgrade Controller through customizable upgrade plans.

Managed Addons

note

System addons are essential for LLMOS to function correctly. However, you can install additional addons beyond the default set as needed.

Creating a Managed Addon​

Managed Addons use Helm for installation and management. You can also add additional addons from external sources like the Artifact Hub.

  1. Navigate to the Managed Addons page and click Create.
  2. Select the desired Namespace and enter a name for the addon.
  3. Provide the Chart Repository URL (e.g., https://grafana.github.io/helm-charts), Chart Name, and Version.
  4. (Optional) Set Enabled to false if you plan to enable the addon later.
  5. (Optional) Customize the addon by adding Values content to override default settings, if necessary.
  6. Click Create to start the installation process.
  7. The addon will install in the background. You can monitor its status on the Managed Addons page.
    • The status will display as Ready once the addon is successfully installed.

Add Managed Addons