kube-cascade/nvidia/README.md

25 lines
916 B
Markdown

# Installing NVIDIA stuff on a new k3s node
curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \
&& curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \
sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \
sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list
sudo apt-get update
sudo apt-get install -y nvidia-container-toolkit nvidia-kernel-dkms
# Installing k3s stuff
There is a `deploy.sh` in this folder.
It installs the RuntimeClass needed to target the nvidia runtime
and installs the device plugin with GFD (node finder for GPUs) via
helm.
# Testing
With these two pieces installed, you should be able to find a GPU-bearing node.
```bash
kubectl get node -l 'nvidia.com/gpu.count'
```