51 lines
2.0 KiB
YAML
51 lines
2.0 KiB
YAML
{{- $Values := (.helm).Values | default .Values }}
|
|
{{- $runbookUrl := ($Values.defaultRules).runbookUrl | default "https://runbooks.prometheus-operator.dev/runbooks" }}
|
|
{{- $clusterLabel := ($Values.global).clusterLabel | default "cluster" }}
|
|
{{- $additionalGroupByLabels := append $Values.defaultRules.additionalGroupByLabels $clusterLabel }}
|
|
{{- $groupLabels := join "," $additionalGroupByLabels }}
|
|
{{- $grafanaHost := ternary (index (($Values.grafana).ingress).hosts 0) (($Values.external).grafana).host ($Values.grafana).enabled }}
|
|
condition: '{{ true }}'
|
|
name: node.rules
|
|
rules:
|
|
- condition: '{{ true }}'
|
|
expr: |-
|
|
topk by (namespace,pod,{{ $groupLabels }}) (1,
|
|
max by (node,namespace,pod,{{ $groupLabels }}) (
|
|
label_replace(kube_pod_info{job="kube-state-metrics",node!=""}, "pod", "$1", "pod", "(.*)")
|
|
))
|
|
record: 'node_namespace_pod:kube_pod_info:'
|
|
- condition: '{{ true }}'
|
|
expr: |-
|
|
count by (node,{{ $groupLabels }}) (
|
|
node_cpu_seconds_total{mode="idle",job="node-exporter"}
|
|
* on (namespace,pod,{{ $groupLabels }}) group_left(node)
|
|
topk by (namespace,pod,{{ $groupLabels }}) (1, node_namespace_pod:kube_pod_info:)
|
|
)
|
|
record: node:node_num_cpu:sum
|
|
- condition: '{{ true }}'
|
|
expr: |-
|
|
sum(
|
|
node_memory_MemAvailable_bytes{job="node-exporter"} or
|
|
(
|
|
node_memory_Buffers_bytes{job="node-exporter"} +
|
|
node_memory_Cached_bytes{job="node-exporter"} +
|
|
node_memory_MemFree_bytes{job="node-exporter"} +
|
|
node_memory_Slab_bytes{job="node-exporter"}
|
|
)
|
|
) by ({{ $groupLabels }})
|
|
record: :node_memory_MemAvailable_bytes:sum
|
|
- condition: '{{ true }}'
|
|
expr: |-
|
|
avg by (node,{{ $groupLabels }}) (
|
|
sum without (mode) (
|
|
rate(node_cpu_seconds_total{mode!="idle",mode!="iowait",mode!="steal",job="node-exporter"}[5m])
|
|
)
|
|
)
|
|
record: node:node_cpu_utilization:ratio_rate5m
|
|
- condition: '{{ true }}'
|
|
expr: |-
|
|
avg by ({{ $groupLabels }}) (
|
|
node:node_cpu_utilization:ratio_rate5m
|
|
)
|
|
record: cluster:node_cpu:ratio_rate5m
|