119 lines
3.5 KiB
YAML
119 lines
3.5 KiB
YAML
apiVersion: v1
|
|
kind: ConfigMap
|
|
metadata:
|
|
name: cluster-health-grafana-dashboard
|
|
namespace: monitoring
|
|
labels:
|
|
grafana_dashboard: "1"
|
|
data:
|
|
cluster-health.json: |
|
|
{
|
|
"title": "Cluster Health / Crash Overview",
|
|
"uid": "cluster-health",
|
|
"schemaVersion": 39,
|
|
"time": { "from": "now-6h", "to": "now" },
|
|
"refresh": "30s",
|
|
"templating": {
|
|
"list": [
|
|
{
|
|
"name": "datasource",
|
|
"type": "datasource",
|
|
"query": "prometheus",
|
|
"current": {}
|
|
}
|
|
]
|
|
},
|
|
"panels": [
|
|
{
|
|
"id": 1,
|
|
"title": "Container aktuell im Waiting-Zustand",
|
|
"type": "stat",
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"gridPos": { "h": 5, "w": 24, "x": 0, "y": 0 },
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{ "color": "green", "value": null },
|
|
{ "color": "red", "value": 1 }
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"targets": [
|
|
{
|
|
"expr": "sum(kube_pod_container_status_waiting == 1) or vector(0)",
|
|
"legendFormat": "Waiting Container",
|
|
"refId": "A"
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"id": 2,
|
|
"title": "Welche Pods/Container warten",
|
|
"type": "table",
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"gridPos": { "h": 8, "w": 24, "x": 0, "y": 5 },
|
|
"targets": [
|
|
{
|
|
"expr": "kube_pod_container_status_waiting == 1",
|
|
"format": "table",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"refId": "A"
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"id": 3,
|
|
"title": "Letzter Terminierungsgrund (ohne 'Completed')",
|
|
"type": "table",
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"gridPos": { "h": 8, "w": 24, "x": 0, "y": 13 },
|
|
"targets": [
|
|
{
|
|
"expr": "kube_pod_container_status_last_terminated_reason{reason!=\"Completed\"} == 1",
|
|
"format": "table",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"refId": "A"
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"id": 4,
|
|
"title": "Pods nicht im Running/Succeeded Zustand",
|
|
"type": "table",
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"gridPos": { "h": 8, "w": 24, "x": 0, "y": 21 },
|
|
"targets": [
|
|
{
|
|
"expr": "kube_pod_status_phase{phase=~\"Failed|Unknown|Pending\"} == 1",
|
|
"format": "table",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"refId": "A"
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"id": 5,
|
|
"title": "Meiste Container-Restarts (letzte Stunde)",
|
|
"type": "table",
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"gridPos": { "h": 8, "w": 24, "x": 0, "y": 29 },
|
|
"targets": [
|
|
{
|
|
"expr": "topk(10, increase(kube_pod_container_status_restarts_total[1h]) > 0)",
|
|
"format": "table",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"refId": "A"
|
|
}
|
|
]
|
|
}
|
|
]
|
|
}
|