Uninstalling the GPU Operator

Perform the following steps to uninstall the Operator.

Optional: List and delete NVIDIA driver custom resources.

$ kubectl get nvidiadrivers

Example Output

NAME          STATUS   AGE
demo-gold     ready    2023-10-16T17:57:12Z
demo-silver   ready    2023-10-16T17:57:12Z

$ kubectl delete nvidiadriver demo-gold
$ kubectl delete nvidiadriver demo-silver

$ kubectl delete crd nvidiadrivers.nvidia.com

Delete the Operator:

$ helm delete -n gpu-operator $(helm list -n gpu-operator | grep gpu-operator | awk '{print $1}')

Optional: List the pods in the Operator namespace to confirm the pods are deleted or in the process of deleting:
```
$ kubectl get pods -n gpu-operator
```
Example Output
```
No resources found.
```

By default, Helm does not support deleting existing CRDs when you delete the chart. As a result, the clusterpolicy CRD and nvidiadrivers CRD will still remain, by default.

$ kubectl get crd clusterpolicies.nvidia.com

To overcome this, the Operator uses a post-delete hook to perform the CRD cleanup. The operator.cleanupCRD chart parameter is added to enable this hook. This parameter is disabled by default. You can enable the hook by specifying --set operator.cleanupCRD=true during install or upgrade to perform automatic CRD cleanup on chart deletion.

Alternatively, you can delete the custom resource definition:

$ kubectl delete crd clusterpolicies.nvidia.com

Note

After uninstalling the Operator, the NVIDIA driver modules might still be loaded. Either reboot the node or unload them using the following command:
```
$ sudo rmmod nvidia_modeset nvidia_uvm nvidia
```
Helm hooks used with the GPU Operator use the Operator image itself. If the Operator image cannot be pulled successfully (either due to network error or an invalid NGC registry secret in case of NVAIE), hooks will fail. In this case, delete the chart and specify the --no-hooks argument to avoid hanging on hook failures.