Intel GPU Metrics
Intel GPU Metrics based on the prometheus metrics from https://github.com/onedr0p/intel-gpu-exporter
This dashboard works for Intel Arc cards or Intel iGPUs, however it has only been configured for one GPU at a time. The objective was to provide a dashboard with a prometheus source, since those available in Grafana labs mostly use InfluxDB.
To configure it correctly, you can use the following docker compose file:
services:
intel-gpu-exporter:
image: ghcr.io/onedr0p/intel-gpu-exporter:rolling
container_name: intel-gpu-exporter
restart: unless-stopped
privileged: true
pid: host
ports:
- "8081:8080"
volumes:
- /dev/dri/:/dev/dri/
And then. in your prometheus scrapers configuration:
# Intel GPU exporter
- job_name: 'intel-gpu' # The job name is hardcoded into the dashboard,
#scrape_interval: 5s
static_configs:
- targets: [ 'host.docker.internal:8081' ]
Data source config
Collector config:
Upload an updated version of an exported dashboard.json file from Grafana
Revision | Description | Created | |
---|---|---|---|
Download |