Enable GPU Usage After Installing DKP
If you want to enable your cluster to run GPU nodes after installing DKP, enable the correct Toolkit version for your operating system in the nvidia-gpu-operator
AppDeployment.
If you have not installed the Kommander component of DKP yet, set the Toolkit version in the Kommander Installer Configuration file and skip this section.
Create a
ConfigMap
with the necessary configuration overrides to set the correct Toolkit version. For example, if you’re using Centos 7.9 or RHEL 7.9 as the base operating system for your GPU enabled nodes, set thetoolkit.version
parameter:CODEcat <<EOF | kubectl apply -f - apiVersion: v1 kind: ConfigMap metadata: namespace: kommander name: nvidia-gpu-operator-overrides data: values.yaml: | toolkit: version: v1.10.0-centos7 EOF
Take the correct GPU Toolkit version from this page: GPU Toolkit Versions
Update the
nvidia-gpu-operator
AppDeployment in thekommander
namespace to reference theConfigMap
you created:CODEcat <<EOF | kubectl apply -f - apiVersion: apps.kommander.d2iq.io/v1alpha3 kind: AppDeployment metadata: name: nvidia-gpu-operator namespace: kommander spec: appRef: name: nvidia-gpu-operator-1.11.1 kind: ClusterApp configOverrides: name: nvidia-gpu-operator-overrides EOF