gpu-operator
gpu-operator copied to clipboard
k8s-driver-manager use a unified version
current in the helm chart values.yaml define many k8s-driver-manager image, but version unlikeness, so we can define a unified field and unified version?
Yes you are right, we do have multiple components which have their own k8s-driver-manager configuration. This could be unified with some refactoring. If you are interested, contributions are welcome on our Gitlab repository (which is where we do development currently): https://gitlab.com/nvidia/kubernetes/gpu-operator