deployments: update to NFD v0.14.0 and drop custom GPU deployment

With the NFD recent versions (v0.13+), it's no longer necessary to
start NFD with custom nfd-master args/rbac settings to get numeric
labels registered as extended resources.

The same can be specified via NodeFeatureRules which also works for
"local" source with feature files.

Signed-off-by: Mikko Ylinen <[email protected]>

Author: mythi

cmd/gpu_plugin/README.md

@@ -227,8 +227,8 @@ take care of, which are explained below. For the RBAC-permissions, gRPC service
 the flag enabling, it is recommended to use kustomization by running:
 
 ```bash
-# Start NFD with GPU related configuration changes
-$ kubectl apply -k 'https://github.com/intel/intel-device-plugins-for-kubernetes/deployments/nfd/overlays/gpu?ref=<RELEASE_VERSION>'
+# Start NFD - if your cluster doesn't have NFD installed yet
+$ kubectl apply -k 'https://github.com/intel/intel-device-plugins-for-kubernetes/deployments/nfd?ref=<RELEASE_VERSION>'
 
 # Create NodeFeatureRules for detecting GPUs on nodes
 $ kubectl apply -k 'https://github.com/intel/intel-device-plugins-for-kubernetes/deployments/nfd/overlays/node-feature-rules?ref=<RELEASE_VERSION>'

deployments/nfd/base/kustomization.yaml

@@ -1,4 +1,9 @@
 apiVersion: kustomize.config.k8s.io/v1beta1
 kind: Kustomization
 resources:
-- "https://github.com/kubernetes-sigs/node-feature-discovery/deployment/overlays/default?ref=v0.13.4"
+- "https://github.com/kubernetes-sigs/node-feature-discovery/deployment/overlays/default?ref=v0.14.0"
+configMapGenerator:
+- name: nfd-worker-conf
+  behavior: replace
+  files:
+  - nfd-worker.conf

deployments/nfd/base/nfd-worker.conf

@@ -0,0 +1,11 @@
+core:
+  featureSources:
+    - "cpu"
+    - "pci"
+    - "local"
+    - "kernel"
+  labelSources:
+   - "-cpu"
+   - "-pci"
+   - "-local"
+   - "-kernel"

deployments/nfd/components/gpu/kustomization.yaml

@@ -4,6 +4,19 @@ metadata:
   name: intel-gpu-platform-labeling
 spec:
   rules:
+  - extendedResources:
+      gpu.intel.com/millicores: "@local.label.gpu.intel.com/millicores"
+      gpu.intel.com/memory.max: "@local.label.gpu.intel.com/memory.max"
+      gpu.intel.com/tiles: "@local.label.gpu.intel.com/tiles"
+    labels:
+      "gpu.intel.com/fractional-resources": "true"
+    matchFeatures:
+      - feature: local.label
+        matchExpressions:
+          gpu.intel.com/millicores: {op: Exists}
+          gpu.intel.com/memory.max: {op: Exists}
+          gpu.intel.com/tiles: {op: Exists}
+    name: intel.gpu.fractionalresources
   # generic rule for older and upcoming devices
   - labels:
     labelsTemplate: |