DirectorySecurity advisories
Sign in

Directory

gpu-feature-discovery logo

gpu-feature-discovery

Last changed
Get started with latest for free
docker pull cgr.dev/chainguard/gpu-feature-discovery

Need access to a specific version? Contact us.

Sign in for updates

Get notified of upcoming product changes, critical vulnerability notifications and patches and more.

Sign in
Versions
Overview
Provenance
Specifications
SBOM
Vulnerabilities
Advisories

Minimal gpu-feature-discovery container image.

Download this Image

The image is available on cgr.dev:

docker pull cgr.dev/chainguard/gpu-feature-discovery:latest

Usage

NVIDIA GPU Feature Discovery for Kubernetes is a software component that allows you to automatically generate labels for the set of GPUs available on a node For more information, refer to the GFD documentation:

NAME:
   GPU Feature Discovery - generate labels for NVIDIA devices

USAGE:
   GPU Feature Discovery [global options] command [command options] [arguments...]

VERSION:
   0.8.2
commit: 47a1ea6862f69844c5364c98a77aa97fe9ea7b74

COMMANDS:
   help, h  Shows a list of commands or help for one command

GLOBAL OPTIONS:
   --mig-strategy value                           the desired strategy for exposing MIG devices on GPUs that support it:
                                                    [none | single | mixed] (default: "none") [$GFD_MIG_STRATEGY, $MIG_STRATEGY]
   --fail-on-init-error                           fail the plugin if an error is encountered during initialization, otherwise block indefinitely (default: true) [$GFD_FAIL_ON_INIT_ERROR, $FAIL_ON_INIT_ERROR]
   --oneshot                                      Label once and exit (default: false) [$GFD_ONESHOT]
   --no-timestamp                                 Do not add the timestamp to the labels (default: false) [$GFD_NO_TIMESTAMP]
   --sleep-interval value                         Time to sleep between labeling (default: 1m0s) [$GFD_SLEEP_INTERVAL]
   --output-file value, --output value, -o value  (default: "/etc/kubernetes/node-feature-discovery/features.d/gfd") [$GFD_OUTPUT_FILE]
   --machine-type-file value                      a path to a file that contains the DMI (SMBIOS) information for the node (default: "/sys/class/dmi/id/product_name") [$GFD_MACHINE_TYPE_FILE]
   --config-file value                            the path to a config file as an alternative to command line options or environment variables [$GFD_CONFIG_FILE, $CONFIG_FILE]
   --use-node-feature-api                         Use NFD NodeFeature API to publish labels (default: false) [$GFD_USE_NODE_FEATURE_API]
   --help, -h                                     show help
   --version, -v                                  print the version

Helm Installation for GPU Feature Discovery

Step 1: Add and Update Helm Repository Add the NVIDIA GFD Helm repository and update it to ensure you have access to the latest charts.

$ helm repo add nvgfd https://nvidia.github.io/gpu-feature-discovery
$ helm repo update

Step 2: Install GPU Feature Discovery Install GPU Feature Discovery using Helm with the specified version, namespace, and optional configuration settings.

This command deploys GPU Feature Discovery as a standalone chart in the gpu-feature-discovery namespace.

$ helm upgrade -i nvgfd nvgfd/gpu-feature-discovery \
  --version 0.8.2 \
  --namespace gpu-feature-discovery \
  --create-namespace

Step 3: Verify Installation Verify that the GPU Feature Discovery DaemonSet is running in the specified namespace.

$ kubectl get daemonset -n gpu-feature-discovery -l app.kubernetes.io/name=gpu-feature-discovery

For more information, refer to the documentation:

Licenses

Chainguard Images contain software packages that are direct or transitive dependencies. The following licenses were found in the "latest" version of this image:

  • Apache-2.0

  • LGPL-2.1-or-later

  • MIT

  • MPL-2.0

For a complete list of licenses, please refer to this Image's SBOM.

Software license agreement

Category
application
tools

Products

Chainguard Images

© 2024 Chainguard, Inc.