titan-iac/services/monitoring/grafana-dashboard-public.yaml

# services/monitoring/grafana-dashboard-public.yaml
apiVersion: v1
kind: ConfigMap
metadata:
  name: grafana-dashboard-public
  labels:
    grafana_dashboard: "1"
data:
  atlas-public-overview.json: |
    {
      "annotations": {
        "list": [
          {
            "builtIn": 1,
            "datasource": {
              "type": "datasource",
              "uid": "grafana"
            },
            "enable": true,
            "hide": true,
            "iconColor": "rgba(0, 211, 255, 1)",
            "name": "Annotations & Alerts",
            "type": "dashboard"
          }
        ]
      },
      "editable": false,
      "fiscalYearStartMonth": 0,
      "graphTooltip": 0,
      "id": null,
      "links": [],
      "liveNow": false,
      "panels": [
        {
          "datasource": {
            "type": "prometheus",
            "uid": "atlas-vm"
          },
          "fieldConfig": {
            "defaults": {
              "color": {
                "mode": "palette-classic"
              },
              "mappings": [],
              "thresholds": {
                "mode": "absolute",
                "steps": [
                  {
                    "color": "green",
                    "value": null
                  }
                ]
              },
              "unit": "none"
            },
            "overrides": []
          },
          "gridPos": {
            "h": 7,
            "w": 6,
            "x": 0,
            "y": 0
          },
          "id": 1,
          "options": {
            "colorMode": "value",
            "graphMode": "area",
            "justifyMode": "auto",
            "orientation": "auto",
            "reduceOptions": {
              "calcs": [
                "lastNotNull"
              ],
              "fields": "",
              "values": false
            },
            "text": {},
            "textMode": "auto"
          },
          "pluginVersion": "10.4.0",
          "targets": [
            {
              "datasource": {
                "type": "prometheus",
                "uid": "atlas-vm"
              },
              "editorMode": "code",
              "expr": "sum(kube_pod_status_phase{phase=\"Running\"})",
              "legendFormat": "",
              "range": true,
              "refId": "A"
            }
          ],
          "title": "Running pods",
          "type": "stat"
        },
        {
          "datasource": {
            "type": "prometheus",
            "uid": "atlas-vm"
          },
          "description": "Aggregated CPU usage across all schedulable nodes.",
          "fieldConfig": {
            "defaults": {
              "color": {
                "mode": "continuous-BlYlRd"
              },
              "mappings": [],
              "max": 100,
              "min": 0,
              "thresholds": {
                "mode": "percentage",
                "steps": [
                  {
                    "color": "green",
                    "value": null
                  },
                  {
                    "color": "yellow",
                    "value": 60
                  },
                  {
                    "color": "red",
                    "value": 85
                  }
                ]
              },
              "unit": "percent"
            },
            "overrides": []
          },
          "gridPos": {
            "h": 7,
            "w": 6,
            "x": 6,
            "y": 0
          },
          "id": 2,
          "options": {
            "colorMode": "value",
            "graphMode": "area",
            "justifyMode": "auto",
            "orientation": "auto",
            "reduceOptions": {
              "calcs": [
                "lastNotNull"
              ],
              "fields": "",
              "values": false
            },
            "text": {},
            "textMode": "auto"
          },
          "targets": [
            {
              "datasource": {
                "type": "prometheus",
                "uid": "atlas-vm"
              },
              "expr": "avg(100 - (avg by (instance) (rate(node_cpu_seconds_total{mode=\"idle\"}[5m])) * 100))",
              "legendFormat": "",
              "refId": "A"
            }
          ],
          "title": "Average node CPU",
          "type": "stat"
        },
        {
          "datasource": {
            "type": "prometheus",
            "uid": "atlas-vm"
          },
          "gridPos": {
            "h": 8,
            "w": 12,
            "x": 0,
            "y": 7
          },
          "id": 3,
          "targets": [
            {
              "datasource": {
                "type": "prometheus",
                "uid": "atlas-vm"
              },
              "expr": "sum(kube_pod_status_phase{phase=\"Running\"}) by (namespace)",
              "legendFormat": "{{namespace}}",
              "refId": "A"
            }
          ],
          "title": "Running pods per namespace",
          "type": "bargauge",
          "options": {
            "displayMode": "gradient",
            "orientation": "horizontal",
            "reduceOptions": {
              "calcs": ["lastNotNull"],
              "fields": "",
              "values": false
            },
            "showUnfilled": false
          }
        },
        {
          "datasource": {
            "type": "prometheus",
            "uid": "atlas-vm"
          },
          "gridPos": {
            "h": 8,
            "w": 12,
            "x": 0,
            "y": 15
          },
          "id": 4,
          "targets": [
            {
              "datasource": {
                "type": "prometheus",
                "uid": "atlas-vm"
              },
              "expr": "sum(kube_node_status_condition{condition=\"Ready\",status=\"true\"})",
              "legendFormat": "Ready",
              "refId": "A"
            },
            {
              "datasource": {
                "type": "prometheus",
                "uid": "atlas-vm"
              },
              "expr": "sum(kube_node_status_condition{condition=\"Ready\",status=\"false\"})",
              "legendFormat": "Not Ready",
              "refId": "B"
            }
          ],
          "title": "Node readiness",
          "type": "piechart",
          "options": {
            "legend": {
              "displayMode": "table",
              "placement": "right"
            },
            "pieType": "donut"
          }
        },
        {
          "datasource": {
            "type": "prometheus",
            "uid": "atlas-vm"
          },
          "gridPos": {
            "h": 8,
            "w": 12,
            "x": 0,
            "y": 23
          },
          "id": 5,
          "targets": [
            {
              "datasource": {
                "type": "prometheus",
                "uid": "atlas-vm"
              },
              "expr": "sum by (namespace) (increase(kube_pod_status_phase{phase=\"Failed\"}[1d]))",
              "legendFormat": "{{namespace}}",
              "refId": "A"
            }
          ],
          "title": "Failed pods (24h)",
          "type": "table",
          "fieldConfig": {
            "defaults": {
              "unit": "none",
              "mappings": [],
              "thresholds": {
                "mode": "absolute",
                "steps": [
                  {"color": "green", "value": null},
                  {"color": "red", "value": 1}
                ]
              }
            },
            "overrides": []
          },
          "options": {
            "showHeader": true
          }
        }
      ],
      "refresh": "30s",
      "schemaVersion": 39,
      "style": "dark",
      "tags": [
        "atlas",
        "public"
      ],
      "templating": {
        "list": []
      },
      "time": {
        "from": "now-12h",
        "to": "now"
      },
      "timepicker": {},
      "timezone": "",
      "title": "Atlas Public Overview",
      "uid": "atlas-public",
      "version": 1,
      "weekStart": ""
    }