This PR refactors the Kueue installation to use the official Helm chart instead of manually maintained, static YAML files. This significantly improves maintainability, simplifies future upgrades, and aligns our toolkit with standard practices.

The previous file-based system was brittle and made updating Kueue a risky, manual process.

Implementation

This change implements a two-stage installation pattern:

Stage 1 (Helm Install): The old file-based install_kueue module is replaced with a new one that uses our generic helm_install module to deploy the official Kueue OCI chart.

Stage 2 (Kubectl Configure): The existing configure_kueue module is preserved to apply our dynamic, blueprint-specific CRDs (ResourceFlavor, ClusterQueue) after the Helm installation completes. A depends_on block was added to ensure correct ordering.

As part of this, the old kueue-*.yaml manifest files and the local variables referencing them have been removed.

Validation

Successfully deployed and validated end-to-end using the gke-a3-ultragpu blueprint.

Helm Release: Verified helm list shows the kueue release as deployed.

Custom Config: Verified kubectl get resourceflavor ,clusterqueue shows the correct, dynamically configured resources.

Performance: a final NCCL benchmark was run successfully, confirming expected hardware performance (~800 GB/s bus bandwidth).

Submission Checklist

NOTE: Community submissions can take up to 2 weeks to be reviewed.

Please take the following actions before submitting this pull request.

• Fork your PR branch from the Toolkit "develop" branch (not main)
• Test all changes with pre-commit in a local branch #
• Confirm that "make tests" passes all tests
• Add or modify unit tests to cover code changes
• Ensure that unit test coverage remains above 80%
• Update all applicable documentation
• Follow Cluster Toolkit Contribution guidelines #