29 lines
1.1 KiB
YAML
29 lines
1.1 KiB
YAML
apiVersion: monitoring.coreos.com/v1
|
|
kind: PrometheusRule
|
|
metadata:
|
|
name: hardware-temperature
|
|
namespace: prometheus
|
|
labels:
|
|
release: prometheus
|
|
spec:
|
|
groups:
|
|
- name: hardware-temperature
|
|
rules:
|
|
- alert: NodeHardwareNearTemperatureLimit
|
|
expr: |-
|
|
max by (instance,chip) (
|
|
(node_hwmon_temp_celsius{job="node-exporter"}
|
|
>= on(instance,chip,sensor)
|
|
(node_hwmon_temp_max_celsius{job="node-exporter"} - 5))
|
|
and on(instance,chip,sensor)
|
|
(node_hwmon_temp_max_celsius{job="node-exporter"} < 200)
|
|
)
|
|
* on(instance,chip) group_left(chip_name) node_hwmon_chip_names{job="node-exporter"}
|
|
* on(instance) group_left(nodename) node_uname_info{job="node-exporter"}
|
|
for: 10m
|
|
labels:
|
|
severity: warning
|
|
annotations:
|
|
summary: "{{ $labels.nodename }} hardware is near its temperature limit"
|
|
description: "{{ $labels.chip_name }} is {{ $value }} C, within 5 C of its reported maximum. Check cooling; thermal throttling can cause CPU, network and storage latency."
|