614 lines
24 KiB
YAML
614 lines
24 KiB
YAML
{{- /*
|
|
Generated from 'alertmanager-overview' from https://raw.githubusercontent.com/prometheus-operator/kube-prometheus/main/manifests/grafana-dashboardDefinitions.yaml
|
|
Do not change in-place! In order to change this file first read following link:
|
|
https://github.com/prometheus-community/helm-charts/tree/main/charts/kube-prometheus-stack/hack
|
|
*/ -}}
|
|
{{- $kubeTargetVersion := default .Capabilities.KubeVersion.GitVersion .Values.kubeTargetVersionOverride }}
|
|
{{- if and (or .Values.grafana.enabled .Values.grafana.forceDeployDashboards) (semverCompare ">=1.14.0-0" $kubeTargetVersion) (semverCompare "<9.9.9-9" $kubeTargetVersion) .Values.grafana.defaultDashboardsEnabled }}
|
|
apiVersion: v1
|
|
kind: ConfigMap
|
|
metadata:
|
|
namespace: {{ template "kube-prometheus-stack-grafana.namespace" . }}
|
|
name: {{ printf "%s-%s" (include "kube-prometheus-stack.fullname" $) "alertmanager-overview" | trunc 63 | trimSuffix "-" }}
|
|
annotations:
|
|
{{ toYaml .Values.grafana.sidecar.dashboards.annotations | indent 4 }}
|
|
labels:
|
|
{{- if $.Values.grafana.sidecar.dashboards.label }}
|
|
{{ $.Values.grafana.sidecar.dashboards.label }}: {{ ternary $.Values.grafana.sidecar.dashboards.labelValue "1" (not (empty $.Values.grafana.sidecar.dashboards.labelValue)) | quote }}
|
|
{{- end }}
|
|
app: {{ template "kube-prometheus-stack.name" $ }}-grafana
|
|
{{ include "kube-prometheus-stack.labels" $ | indent 4 }}
|
|
data:
|
|
alertmanager-overview.json: |-
|
|
{
|
|
"__inputs": [
|
|
|
|
],
|
|
"__requires": [
|
|
|
|
],
|
|
"annotations": {
|
|
"list": [
|
|
|
|
]
|
|
},
|
|
"editable": false,
|
|
"gnetId": null,
|
|
"graphTooltip": 1,
|
|
"hideControls": false,
|
|
"id": null,
|
|
"links": [
|
|
|
|
],
|
|
"refresh": "30s",
|
|
"rows": [
|
|
{
|
|
"collapse": false,
|
|
"collapsed": false,
|
|
"panels": [
|
|
{
|
|
"aliasColors": {
|
|
|
|
},
|
|
"bars": false,
|
|
"dashLength": 10,
|
|
"dashes": false,
|
|
"datasource": "$datasource",
|
|
"description": "current set of alerts stored in the Alertmanager",
|
|
"fill": 1,
|
|
"fillGradient": 0,
|
|
"gridPos": {
|
|
|
|
},
|
|
"id": 2,
|
|
"legend": {
|
|
"alignAsTable": false,
|
|
"avg": false,
|
|
"current": false,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": false,
|
|
"show": false,
|
|
"sideWidth": null,
|
|
"total": false,
|
|
"values": false
|
|
},
|
|
"lines": true,
|
|
"linewidth": 1,
|
|
"links": [
|
|
|
|
],
|
|
"nullPointMode": "null",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"repeat": null,
|
|
"seriesOverrides": [
|
|
|
|
],
|
|
"spaceLength": 10,
|
|
"span": 6,
|
|
"stack": true,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "sum(alertmanager_alerts{namespace=~\"$namespace\",service=~\"$service\"}) by (namespace,service,instance)",
|
|
"format": "time_series",
|
|
"intervalFactor": 2,
|
|
"legendFormat": "{{`{{`}}instance{{`}}`}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"thresholds": [
|
|
|
|
],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "Alerts",
|
|
"tooltip": {
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "individual"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"buckets": null,
|
|
"mode": "time",
|
|
"name": null,
|
|
"show": true,
|
|
"values": [
|
|
|
|
]
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "none",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "none",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"aliasColors": {
|
|
|
|
},
|
|
"bars": false,
|
|
"dashLength": 10,
|
|
"dashes": false,
|
|
"datasource": "$datasource",
|
|
"description": "rate of successful and invalid alerts received by the Alertmanager",
|
|
"fill": 1,
|
|
"fillGradient": 0,
|
|
"gridPos": {
|
|
|
|
},
|
|
"id": 3,
|
|
"legend": {
|
|
"alignAsTable": false,
|
|
"avg": false,
|
|
"current": false,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": false,
|
|
"show": false,
|
|
"sideWidth": null,
|
|
"total": false,
|
|
"values": false
|
|
},
|
|
"lines": true,
|
|
"linewidth": 1,
|
|
"links": [
|
|
|
|
],
|
|
"nullPointMode": "null",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"repeat": null,
|
|
"seriesOverrides": [
|
|
|
|
],
|
|
"spaceLength": 10,
|
|
"span": 6,
|
|
"stack": true,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "sum(rate(alertmanager_alerts_received_total{namespace=~\"$namespace\",service=~\"$service\"}[$__rate_interval])) by (namespace,service,instance)",
|
|
"format": "time_series",
|
|
"intervalFactor": 2,
|
|
"legendFormat": "{{`{{`}}instance{{`}}`}} Received",
|
|
"refId": "A"
|
|
},
|
|
{
|
|
"expr": "sum(rate(alertmanager_alerts_invalid_total{namespace=~\"$namespace\",service=~\"$service\"}[$__rate_interval])) by (namespace,service,instance)",
|
|
"format": "time_series",
|
|
"intervalFactor": 2,
|
|
"legendFormat": "{{`{{`}}instance{{`}}`}} Invalid",
|
|
"refId": "B"
|
|
}
|
|
],
|
|
"thresholds": [
|
|
|
|
],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "Alerts receive rate",
|
|
"tooltip": {
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "individual"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"buckets": null,
|
|
"mode": "time",
|
|
"name": null,
|
|
"show": true,
|
|
"values": [
|
|
|
|
]
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "ops",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "ops",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
}
|
|
]
|
|
}
|
|
],
|
|
"repeat": null,
|
|
"repeatIteration": null,
|
|
"repeatRowId": null,
|
|
"showTitle": true,
|
|
"title": "Alerts",
|
|
"titleSize": "h6",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"collapse": false,
|
|
"collapsed": false,
|
|
"panels": [
|
|
{
|
|
"aliasColors": {
|
|
|
|
},
|
|
"bars": false,
|
|
"dashLength": 10,
|
|
"dashes": false,
|
|
"datasource": "$datasource",
|
|
"description": "rate of successful and invalid notifications sent by the Alertmanager",
|
|
"fill": 1,
|
|
"fillGradient": 0,
|
|
"gridPos": {
|
|
|
|
},
|
|
"id": 4,
|
|
"legend": {
|
|
"alignAsTable": false,
|
|
"avg": false,
|
|
"current": false,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": false,
|
|
"show": false,
|
|
"sideWidth": null,
|
|
"total": false,
|
|
"values": false
|
|
},
|
|
"lines": true,
|
|
"linewidth": 1,
|
|
"links": [
|
|
|
|
],
|
|
"nullPointMode": "null",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"repeat": "integration",
|
|
"seriesOverrides": [
|
|
|
|
],
|
|
"spaceLength": 10,
|
|
"stack": true,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "sum(rate(alertmanager_notifications_total{namespace=~\"$namespace\",service=~\"$service\", integration=\"$integration\"}[$__rate_interval])) by (integration,namespace,service,instance)",
|
|
"format": "time_series",
|
|
"intervalFactor": 2,
|
|
"legendFormat": "{{`{{`}}instance{{`}}`}} Total",
|
|
"refId": "A"
|
|
},
|
|
{
|
|
"expr": "sum(rate(alertmanager_notifications_failed_total{namespace=~\"$namespace\",service=~\"$service\", integration=\"$integration\"}[$__rate_interval])) by (integration,namespace,service,instance)",
|
|
"format": "time_series",
|
|
"intervalFactor": 2,
|
|
"legendFormat": "{{`{{`}}instance{{`}}`}} Failed",
|
|
"refId": "B"
|
|
}
|
|
],
|
|
"thresholds": [
|
|
|
|
],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "$integration: Notifications Send Rate",
|
|
"tooltip": {
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "individual"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"buckets": null,
|
|
"mode": "time",
|
|
"name": null,
|
|
"show": true,
|
|
"values": [
|
|
|
|
]
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "ops",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "ops",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"aliasColors": {
|
|
|
|
},
|
|
"bars": false,
|
|
"dashLength": 10,
|
|
"dashes": false,
|
|
"datasource": "$datasource",
|
|
"description": "latency of notifications sent by the Alertmanager",
|
|
"fill": 1,
|
|
"fillGradient": 0,
|
|
"gridPos": {
|
|
|
|
},
|
|
"id": 5,
|
|
"legend": {
|
|
"alignAsTable": false,
|
|
"avg": false,
|
|
"current": false,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": false,
|
|
"show": false,
|
|
"sideWidth": null,
|
|
"total": false,
|
|
"values": false
|
|
},
|
|
"lines": true,
|
|
"linewidth": 1,
|
|
"links": [
|
|
|
|
],
|
|
"nullPointMode": "null",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"repeat": "integration",
|
|
"seriesOverrides": [
|
|
|
|
],
|
|
"spaceLength": 10,
|
|
"stack": false,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "histogram_quantile(0.99,\n sum(rate(alertmanager_notification_latency_seconds_bucket{namespace=~\"$namespace\",service=~\"$service\", integration=\"$integration\"}[$__rate_interval])) by (le,namespace,service,instance)\n) \n",
|
|
"format": "time_series",
|
|
"intervalFactor": 2,
|
|
"legendFormat": "{{`{{`}}instance{{`}}`}} 99th Percentile",
|
|
"refId": "A"
|
|
},
|
|
{
|
|
"expr": "histogram_quantile(0.50,\n sum(rate(alertmanager_notification_latency_seconds_bucket{namespace=~\"$namespace\",service=~\"$service\", integration=\"$integration\"}[$__rate_interval])) by (le,namespace,service,instance)\n) \n",
|
|
"format": "time_series",
|
|
"intervalFactor": 2,
|
|
"legendFormat": "{{`{{`}}instance{{`}}`}} Median",
|
|
"refId": "B"
|
|
},
|
|
{
|
|
"expr": "sum(rate(alertmanager_notification_latency_seconds_sum{namespace=~\"$namespace\",service=~\"$service\", integration=\"$integration\"}[$__rate_interval])) by (namespace,service,instance)\n/\nsum(rate(alertmanager_notification_latency_seconds_count{namespace=~\"$namespace\",service=~\"$service\", integration=\"$integration\"}[$__rate_interval])) by (namespace,service,instance)\n",
|
|
"format": "time_series",
|
|
"intervalFactor": 2,
|
|
"legendFormat": "{{`{{`}}instance{{`}}`}} Average",
|
|
"refId": "C"
|
|
}
|
|
],
|
|
"thresholds": [
|
|
|
|
],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "$integration: Notification Duration",
|
|
"tooltip": {
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "individual"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"buckets": null,
|
|
"mode": "time",
|
|
"name": null,
|
|
"show": true,
|
|
"values": [
|
|
|
|
]
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "s",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "s",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
}
|
|
]
|
|
}
|
|
],
|
|
"repeat": null,
|
|
"repeatIteration": null,
|
|
"repeatRowId": null,
|
|
"showTitle": true,
|
|
"title": "Notifications",
|
|
"titleSize": "h6",
|
|
"type": "row"
|
|
}
|
|
],
|
|
"schemaVersion": 14,
|
|
"style": "dark",
|
|
"tags": [
|
|
"alertmanager-mixin"
|
|
],
|
|
"templating": {
|
|
"list": [
|
|
{
|
|
"current": {
|
|
"text": "Prometheus",
|
|
"value": "Prometheus"
|
|
},
|
|
"hide": 0,
|
|
"label": "Data Source",
|
|
"name": "datasource",
|
|
"options": [
|
|
|
|
],
|
|
"query": "prometheus",
|
|
"refresh": 1,
|
|
"regex": "",
|
|
"type": "datasource"
|
|
},
|
|
{
|
|
"allValue": null,
|
|
"current": {
|
|
"text": "",
|
|
"value": ""
|
|
},
|
|
"datasource": "$datasource",
|
|
"hide": 0,
|
|
"includeAll": false,
|
|
"label": "namespace",
|
|
"multi": false,
|
|
"name": "namespace",
|
|
"options": [
|
|
|
|
],
|
|
"query": "label_values(alertmanager_alerts, namespace)",
|
|
"refresh": 2,
|
|
"regex": "",
|
|
"sort": 1,
|
|
"tagValuesQuery": "",
|
|
"tags": [
|
|
|
|
],
|
|
"tagsQuery": "",
|
|
"type": "query",
|
|
"useTags": false
|
|
},
|
|
{
|
|
"allValue": null,
|
|
"current": {
|
|
"text": "",
|
|
"value": ""
|
|
},
|
|
"datasource": "$datasource",
|
|
"hide": 0,
|
|
"includeAll": false,
|
|
"label": "service",
|
|
"multi": false,
|
|
"name": "service",
|
|
"options": [
|
|
|
|
],
|
|
"query": "label_values(alertmanager_alerts, service)",
|
|
"refresh": 2,
|
|
"regex": "",
|
|
"sort": 1,
|
|
"tagValuesQuery": "",
|
|
"tags": [
|
|
|
|
],
|
|
"tagsQuery": "",
|
|
"type": "query",
|
|
"useTags": false
|
|
},
|
|
{
|
|
"allValue": null,
|
|
"current": {
|
|
"text": "all",
|
|
"value": "$__all"
|
|
},
|
|
"datasource": "$datasource",
|
|
"hide": 2,
|
|
"includeAll": true,
|
|
"label": null,
|
|
"multi": false,
|
|
"name": "integration",
|
|
"options": [
|
|
|
|
],
|
|
"query": "label_values(alertmanager_notifications_total{integration=~\".*\"}, integration)",
|
|
"refresh": 2,
|
|
"regex": "",
|
|
"sort": 1,
|
|
"tagValuesQuery": "",
|
|
"tags": [
|
|
|
|
],
|
|
"tagsQuery": "",
|
|
"type": "query",
|
|
"useTags": false
|
|
}
|
|
]
|
|
},
|
|
"time": {
|
|
"from": "now-1h",
|
|
"to": "now"
|
|
},
|
|
"timepicker": {
|
|
"refresh_intervals": [
|
|
"5s",
|
|
"10s",
|
|
"30s",
|
|
"1m",
|
|
"5m",
|
|
"15m",
|
|
"30m",
|
|
"1h",
|
|
"2h",
|
|
"1d"
|
|
],
|
|
"time_options": [
|
|
"5m",
|
|
"15m",
|
|
"1h",
|
|
"6h",
|
|
"12h",
|
|
"24h",
|
|
"2d",
|
|
"7d",
|
|
"30d"
|
|
]
|
|
},
|
|
"timezone": "{{ .Values.grafana.defaultDashboardsTimezone }}",
|
|
"title": "Alertmanager / Overview",
|
|
"uid": "alertmanager-overview",
|
|
"version": 0
|
|
}
|
|
{{- end }} |