chore(monitoring): tidy GPU VMRule top-level structure

Drop the hardcoded metadata.namespace so the rule inherits the chart's
release namespace, and add an explicit empty params field on every
group for schema consistency. No behavior change.

Signed-off-by: Arsolitt <arsolitt@gmail.com>
This commit is contained in:
Arsolitt 2026-04-18 07:45:02 +03:00
parent b64bfcc414
commit 51b0dedd08
No known key found for this signature in database
GPG key ID: 4D8302CE6A9247C4

View file

@ -2,11 +2,11 @@ apiVersion: operator.victoriametrics.com/v1beta1
kind: VMRule
metadata:
name: alerts-gpu-recording.rules
namespace: cozy-monitoring
spec:
groups:
- name: gpu.recording.cluster.1m
interval: 1m
params: {}
rules:
- record: cluster:gpu_count:total
expr: count(group by (UUID) (DCGM_FI_DEV_GPU_UTIL))
@ -25,6 +25,7 @@ spec:
- name: gpu.recording.namespace.1m
interval: 1m
params: {}
rules:
# Kube-requested GPU count per namespace — billable view, includes
# Pending pods. Use this for GPU-hour reporting via
@ -61,6 +62,7 @@ spec:
- name: gpu.recording.efficiency.1m
interval: 1m
params: {}
rules:
# Tensor hardware saturation — the honest "am I using the GPU" metric
# for AI/LLM workloads. Unlike NVML, idle tensor cores are visible.