Skip to content

Instantly share code, notes, and snippets.

@fenar

fenar/NVIDIA-GPU-Observability.md

Last active June 12, 2024 22:02
Show Gist options
  • Save fenar/081b22237b5eaa5398dafa70a80fdff1 to your computer and use it in GitHub Desktop.
Save fenar/081b22237b5eaa5398dafa70a80fdff1 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
NVIDIA-GPU-Observability.md

Configure the NVIDIA DCGM Exporter Dashboard on Red Hat OCP

(1) Download the latest NVIDIA DCGM Exporter Dashboard from the DCGM Exporter repository on GitHub:
curl -LfO https://github.com/NVIDIA/dcgm-exporter/raw/main/grafana/dcgm-exporter-dashboard.json

(2) Create a config map from the downloaded file in the openshift-config-managed namespace:
oc create configmap nvidia-dcgm-exporter-dashboard -n openshift-config-managed --from-file=dcgm-exporter-dashboard.json

(3) Label the config map to expose the dashboard in the Administrator perspective of the web console:
oc label configmap nvidia-dcgm-exporter-dashboard -n openshift-config-managed "console.openshift.io/dashboard=true"

(4) Label the config map to expose the dashboard in the Developer perspecitive of the web console:
oc label configmap nvidia-dcgm-exporter-dashboard -n openshift-config-managed "console.openshift.io/odc-dashboard=true"

(5) View the created resource and verify the labels:
oc -n openshift-config-managed get cm nvidia-dcgm-exporter-dashboard --show-labels

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment