Merge pull request Mellanox#557 from ArangoGutierrez/nfd_nfr

Sync NFD Helm Charts and config with the GPU-Operator

Dockerfile

@@ -40,7 +40,8 @@ COPY deployment/network-operator chart
 # copy CRDs from helm charts
 RUN mkdir crds && \
     cp -r chart/crds /workspace/crds/network-operator/ && \
-    cp -r chart/charts/sriov-network-operator/crds /workspace/crds/sriov-network-operator/
+    cp -r chart/charts/sriov-network-operator/crds /workspace/crds/sriov-network-operator/ && \
+    cp -r chart/charts/node-feature-discovery/crds /workspace/crds/node-feature-discovery/
 
 # Build
 RUN CGO_ENABLED=0 GOOS=linux GOARCH=amd64 go build -a -o manager main.go

README.md

@@ -50,7 +50,8 @@ For more information please visit the official [documentation](https://docs.nvid
 ## Prerequisites
 ### Kubernetes Node Feature Discovery (NFD)
 NVIDIA Network operator relies on Node labeling to get the cluster to the desired state.
-[Node Feature Discovery](https://github.com/kubernetes-sigs/node-feature-discovery) `v0.10.1` or newer is expected to be deployed to provide the appropriate labeling:
+[Node Feature Discovery](https://github.com/kubernetes-sigs/node-feature-discovery) `v0.13.2` or newer is deployed by default via HELM chart installation.
+NFD is used to label nodes with the following labels:
 
 - PCI vendor and device information
 - RDMA capability
@@ -59,19 +60,23 @@ NVIDIA Network operator relies on Node labeling to get the cluster to the desire
 __Example NFD worker configurations:__
 
 ```yaml
-sources:
-  custom:
-  pci:
-    deviceClassWhitelist:
-      - "02"
-      - "0200"
-      - "0207"
-    deviceLabelFields:
-      - "vendor"
+    config:
+      sources:
+        pci:
+          deviceClassWhitelist:
+          - "02"
+          - "0200"
+          - "0207"
+          - "0300"
+          - "0302"
+          deviceLabelFields:
+          - vendor
 ```
 
 >\* Required for GPUDirect driver container deployment
 
+If NFD is already deployed in the cluster, make sure to pass `--set nfd.enabled=false` to the helm install command to avoid conflicts.
+
 ## Resource Definitions
 The Operator Acts on the following CRDs:
 

deployment/network-operator/Chart.yaml

@@ -15,7 +15,7 @@ dependencies:
 - condition: nfd.enabled
   name: node-feature-discovery
   repository: "http://kubernetes-sigs.github.io/node-feature-discovery/charts"
-  version: 0.10.1
+  version: 0.13.2
 - condition: sriovNetworkOperator.enabled
   name: sriov-network-operator
   repository: ""

deployment/network-operator/charts/node-feature-discovery/.helmignore

@@ -0,0 +1,23 @@
+# Patterns to ignore when building packages.
+# This supports shell glob matching, relative path matching, and
+# negation (prefixed with !). Only one pattern per line.
+.DS_Store
+# Common VCS dirs
+.git/
+.gitignore
+.bzr/
+.bzrignore
+.hg/
+.hgignore
+.svn/
+# Common backup files
+*.swp
+*.bak
+*.tmp
+*.orig
+*~
+# Various IDEs
+.project
+.idea/
+*.tmproj
+.vscode/

deployment/network-operator/charts/node-feature-discovery/Chart.yaml

@@ -0,0 +1,14 @@
+apiVersion: v2
+appVersion: v0.13.2
+description: 'Detects hardware features available on each node in a Kubernetes cluster,
+  and advertises those features using node labels. '
+home: https://github.com/kubernetes-sigs/node-feature-discovery
+keywords:
+- feature-discovery
+- feature-detection
+- node-labels
+name: node-feature-discovery
+sources:
+- https://github.com/kubernetes-sigs/node-feature-discovery
+type: application
+version: 0.13.2

deployment/network-operator/charts/node-feature-discovery/README.md

@@ -0,0 +1,10 @@
+# Node Feature Discovery
+
+Node Feature Discovery (NFD) is a Kubernetes add-on for detecting hardware
+features and system configuration. Detected features are advertised as node
+labels. NFD provides flexible configuration and extension points for a wide
+range of vendor and application specific node labeling needs.
+
+See
+[NFD documentation](https://kubernetes-sigs.github.io/node-feature-discovery/v0.13/deployment/helm.html)
+for deployment instructions.