Skip to content

Conversation

@claudia-lola
Copy link
Contributor

This pull request updates the CUDA Ansible role defaults to improve flexibility and control over package installation, particularly for the NVIDIA Fabric Manager. The main changes introduce a new option to conditionally install the Fabric Manager and refactor how package lists are constructed.

Improvements to package management:

  • Added a new boolean option cuda_install_nvidiafabricmanger to control whether the NVIDIA Fabric Manager is installed, defaulting to false.
  • Split the CUDA package list into cuda_packages_default and cuda_packages_fabricmanager, and refactored cuda_packages to dynamically include the Fabric Manager package based on the new option.

Versioning and configuration updates:

  • Introduced cuda_nvidia_driver_version variable for clearer driver version management and updated cuda_nvidia_driver_pkg to use it.
  • Added cuda_packages_fabricmanager to specify the Fabric Manager package version using the new driver version variable.

@claudia-lola claudia-lola requested a review from a team as a code owner October 23, 2025 12:32
Copy link
Collaborator

@sjpb sjpb left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Once changes are done, could you manually trigger an extra build on this branch please, AND add a comment linking to the workflow run?

Then we'll merge once that build passes/comments are addressed.

Remove commented option for installing nvidia-fabricmanager.
@claudia-lola
Copy link
Contributor Author

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants