- nic-feature-discovery
NVIDIA NIC Feature Discovery for Kubernetes is a software component that allows you to automatically generate Node labels for NIC related features available on a K8s Node. It leverages the Node Feature Discovery to perform this labeling.
- Kubernetes >=
1.24 - Node Feature Discovery (NFD) >=
0.13.2- Deployed on each node where you want to label with the local source configured
- To deploy NFD, refer to the project Official Documentation
| Label Name | Value Type | Description | Example |
|---|---|---|---|
| nvidia.com/mofed.version | String | MOFED driver version if present and loaded | "23.04-0.5.3" |
Refer to Node Feature Discovery - Quick Start
Example deployment using kustomize:
kubectl apply -k https://github.com/kubernetes-sigs/node-feature-discovery/deployment/overlays/default?ref=v0.13.3$ kubectl apply -k https://github.com/Mellanox/nic-feature-discovery/deployment/k8s/overlays/default?ref=main- Node Feature Discovery is deployed
root:~# kubectl get -n node-feature-discovery all
NAME READY STATUS RESTARTS AGE
pod/nfd-master-5c56499456-r7g2h 1/1 Running 0 3d23h
pod/nfd-worker-9qbdh 1/1 Running 6 (16h ago) 3d23h
pod/nfd-worker-w6twz 1/1 Running 0 3d23h
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
service/nfd-master ClusterIP 10.96.9.245 <none> 8080/TCP 3d23h
NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE
daemonset.apps/nfd-worker 2 2 2 2 2 <none> 3d23h
NAME READY UP-TO-DATE AVAILABLE AGE
deployment.apps/nfd-master 1/1 1 1 3d23h
NAME DESIRED CURRENT READY AGE
replicaset.apps/nfd-master-5c56499456 1 1 1 3d23h- NVIDIA NIC Feature discovery is deployed
root:~# kubectl get -n nic-feature-discovery all
NAME READY STATUS RESTARTS AGE
pod/nic-feature-discovery-ds-ln4r9 1/1 Running 0 3d23h
pod/nic-feature-discovery-ds-rtvjj 1/1 Running 0 3d23h
pod/nic-feature-discovery-ds-tbtr2 1/1 Running 0 3d23h
NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE
daemonset.apps/nic-feature-discovery-ds 3 3 3 3 3 <none> 3d23h- Node contain expected labels
root:~# kubectl describe node my-worker-node
Name: my-worker-node
Roles: <none>
Labels: beta.kubernetes.io/arch=amd64
beta.kubernetes.io/os=linux
feature.node.kubernetes.io/cpu-cpuid.AESNI=true
feature.node.kubernetes.io/cpu-cpuid.AVX=true
feature.node.kubernetes.io/cpu-cpuid.CMPXCHG8=true
feature.node.kubernetes.io/cpu-cpuid.FLUSH_L1D=true
feature.node.kubernetes.io/cpu-cpuid.FXSR=true
feature.node.kubernetes.io/cpu-cpuid.FXSROPT=true
feature.node.kubernetes.io/cpu-cpuid.IBPB=true
....
...
..
.
nvidia.com/mofed.version=23.04-0.5.3Usage:
nic-feature-discovery [flags]
Feature Discovery Daemon flags:
--features-file-name string
features file name (default "nvidia-com-nic-feature-discovery.features")
--features-scan-interval duration
features scan interval (default 1m0s)
--nfd-features-path string
node feature discovery local features path (default "/etc/kubernetes/node-feature-discovery/features.d/")
Logging flags:
--log-flush-frequency duration
Maximum number of seconds between log flushes (default 5s)
--log-json-info-buffer-size quantity
[Alpha] In JSON format with split output streams, the info messages can be buffered for a while to increase performance. The default value of zero bytes disables buffering. The size can be specified as number of bytes (512),
multiples of 1000 (1K), multiples of 1024 (2Ki), or powers of those (3M, 4G, 5Mi, 6Gi). Enable the LoggingAlphaOptions feature gate to use this.
--log-json-split-stream
[Alpha] In JSON format, write error messages to stderr and info messages to stdout. The default is to write a single stream to stdout. Enable the LoggingAlphaOptions feature gate to use this.
--logging-format string
Sets the log format. Permitted formats: "json" (gated by LoggingBetaOptions), "text". (default "text")
-v, --v Level
number for the log level verbosity
--vmodule pattern=N,...
comma-separated list of pattern=N settings for file-filtered logging (only works for text log format)
General flags:
-h, --help
print help and exit
--version
print version and exit
- The general prerequitites
- golang >= 1.20
- install Task
go install github.com/go-task/task/v3/cmd/task@latest- clone project
git clone https://github.com/Mellanox/nic-feature-discovery.git- build binary
cd nic-feature-discovery
task build- run binary
./build/nic-feature-discoveryNote: To build container image run
task image:build. to deploy this image in your k8 cluster, you should re-tag and upload to your own image registry, then create an overlay for exsiting deployment which overrides the image name with your own image path.
For development skaffold is used to deploy local changes onto a K8s cluster.
task -o interleaved localdev:startTo trigger a rebuild, press any key in the current shell.
task -o interleaved localdev:cleanFor this deployment option you should provide the following environment variables
LOCALDEV_IMAGE_REPO: image repository which skaffold will use.LOCALDEV_KUBECONFIG: optional, path to kubeconfig file which will be used by skaffold to access k8s cluster.LOCALDEV_KUBECONTEXT: optional. name of kube context which will be used by skaffold to access k8s cluster.
LOCALDEV_IMAGE_REPO=quay.io/myuser LOCALDEV_KUBECONTEXT=my-remote-cluster task -o interleaved localdev:deployTo trigger a rebuild, press any key in the current shell.
NOTE: if
LOCALDEV_KUBECONFIGandLOCALDEV_KUBECONTEXTare not provided, the current context pointed by kubectl will be used.