vrr/cozystack
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fix(fleet): guard TDP division and document DCGM dependency

Browse source 
Signed-off-by: Arsolitt <arsolitt@gmail.com>
This commit is contained in:
Arsolitt 2026-04-19 11:01:40 +03:00
parent 14d9188fcd
commit 950c5dd669
No known key found for this signature in database
GPG key ID: 4D8302CE6A9247C4
1 changed files with 2 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

4
dashboards/gpu/gpu-fleet.json
View file

@ -467,13 +467,13 @@
"id": 12,
"targets": [
{
"expr": "sum by (Hostname) (DCGM_FI_DEV_POWER_USAGE) / sum by (Hostname) (DCGM_FI_DEV_POWER_MGMT_LIMIT) * 100",
"expr": "sum by (Hostname) (DCGM_FI_DEV_POWER_USAGE) / clamp_min(sum by (Hostname) (DCGM_FI_DEV_POWER_MGMT_LIMIT), 1) * 100",
"legendFormat": "{{Hostname}}",
"refId": "A"
}
],
"title": "Power draw per node (% of TDP)",
"description": "GPU power usage as a fraction of the node's combined TDP cap. Works across GPU SKUs without per-model thresholds.",
"description": "GPU power usage as a fraction of the node's combined TDP cap. Works across GPU SKUs without per-model thresholds. Requires DCGM_FI_DEV_POWER_MGMT_LIMIT from custom DCGM CSV (see packages/system/gpu-operator/examples/dcgm-custom-metrics.yaml).",
"transparent": false,
"datasource": {
"type": "prometheus",