fix: use nvidia runtimeClassName for device plugin GPU access

2026-03-01 17:56:16 +02:00 · 2026-03-01 17:56:16 +02:00 · ef9c258a55
commit ef9c258a55
parent a34002aa0f
1 changed files with 2 additions and 0 deletions
--- a/manifests/nvidia-device-plugin/nvidia-device-plugin.yaml
+++ b/manifests/nvidia-device-plugin/nvidia-device-plugin.yaml
@ -46,6 +46,8 @@ spec:
      # be rescheduled after a failure.
      # See https://kubernetes.io/docs/tasks/administer-cluster/guaranteed-scheduling-critical-addon-pods/
      priorityClassName: "system-node-critical"
+      # Use NVIDIA runtime to provide GPU device access
+      runtimeClassName: nvidia
      containers:
      - image: nvcr.io/nvidia/k8s-device-plugin:v0.15.0
        name: nvidia-device-plugin-ctr