12cd0fd34f
This also contains the generated ConfigMaps for the grafana dashboards as well as for alerts. These will be updated in subsequent changes as sets of alerts and dashboards are added/updated in the respective subdirectories in `/assets`.
589 lines
23 KiB
JSON
589 lines
23 KiB
JSON
{
|
|
"dashboard": {
|
|
"__inputs": [
|
|
{
|
|
"description": "",
|
|
"label": "prometheus",
|
|
"name": "DS_PROMETHEUS",
|
|
"pluginId": "prometheus",
|
|
"pluginName": "Prometheus",
|
|
"type": "datasource"
|
|
}
|
|
],
|
|
"__requires": [
|
|
{
|
|
"id": "singlestat",
|
|
"name": "Singlestat",
|
|
"type": "panel",
|
|
"version": ""
|
|
},
|
|
{
|
|
"id": "graph",
|
|
"name": "Graph",
|
|
"type": "panel",
|
|
"version": ""
|
|
},
|
|
{
|
|
"id": "grafana",
|
|
"name": "Grafana",
|
|
"type": "grafana",
|
|
"version": "3.1.1"
|
|
},
|
|
{
|
|
"id": "prometheus",
|
|
"name": "Prometheus",
|
|
"type": "datasource",
|
|
"version": "1.0.0"
|
|
}
|
|
],
|
|
"annotations": {
|
|
"list": []
|
|
},
|
|
"description": "",
|
|
"editable": true,
|
|
"gnetId": 162,
|
|
"hideControls": true,
|
|
"id": null,
|
|
"links": [
|
|
{
|
|
"icon": "external link",
|
|
"includeVars": true,
|
|
"tags": [],
|
|
"targetBlank": false,
|
|
"type": "dashboards"
|
|
}
|
|
],
|
|
"refresh": "10s",
|
|
"rows": [
|
|
{
|
|
"collapse": false,
|
|
"editable": true,
|
|
"height": "250px",
|
|
"panels": [
|
|
{
|
|
"cacheTimeout": null,
|
|
"colorBackground": false,
|
|
"colorValue": false,
|
|
"colors": [
|
|
"rgba(50, 172, 45, 0.97)",
|
|
"rgba(237, 129, 40, 0.89)",
|
|
"rgba(245, 54, 54, 0.9)"
|
|
],
|
|
"datasource": "${DS_PROMETHEUS}",
|
|
"decimals": 2,
|
|
"editable": true,
|
|
"error": false,
|
|
"format": "percent",
|
|
"gauge": {
|
|
"maxValue": 100,
|
|
"minValue": 0,
|
|
"show": true,
|
|
"thresholdLabels": false,
|
|
"thresholdMarkers": true
|
|
},
|
|
"id": 6,
|
|
"interval": null,
|
|
"isNew": true,
|
|
"links": [],
|
|
"mappingType": 1,
|
|
"mappingTypes": [
|
|
{
|
|
"name": "value to text",
|
|
"value": 1
|
|
},
|
|
{
|
|
"name": "range to text",
|
|
"value": 2
|
|
}
|
|
],
|
|
"maxDataPoints": 100,
|
|
"nullPointMode": "connected",
|
|
"nullText": null,
|
|
"postfix": "",
|
|
"postfixFontSize": "50%",
|
|
"prefix": "",
|
|
"prefixFontSize": "50%",
|
|
"rangeMaps": [
|
|
{
|
|
"from": "null",
|
|
"text": "N/A",
|
|
"to": "null"
|
|
}
|
|
],
|
|
"span": 3,
|
|
"sparkline": {
|
|
"fillColor": "rgba(31, 118, 189, 0.18)",
|
|
"full": false,
|
|
"lineColor": "rgb(31, 120, 193)",
|
|
"show": false
|
|
},
|
|
"targets": [
|
|
{
|
|
"expr": "sum(sum by (container_name)( rate(container_cpu_usage_seconds_total{image!=\"\",instance=~\"$node\"}[1m] ) )) / count(node_cpu{mode=\"system\",instance=~\"$node:9100\"}) * 100",
|
|
"interval": "10s",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "",
|
|
"refId": "A",
|
|
"step": 10
|
|
}
|
|
],
|
|
"thresholds": "65, 90",
|
|
"title": "Cluster CPU Usage",
|
|
"type": "singlestat",
|
|
"valueFontSize": "80%",
|
|
"valueMaps": [
|
|
{
|
|
"op": "=",
|
|
"text": "N/A",
|
|
"value": "null"
|
|
}
|
|
],
|
|
"valueName": "current"
|
|
},
|
|
{
|
|
"aliasColors": {},
|
|
"bars": false,
|
|
"datasource": "${DS_PROMETHEUS}",
|
|
"decimals": 3,
|
|
"editable": true,
|
|
"error": false,
|
|
"fill": 0,
|
|
"grid": {
|
|
"threshold1": null,
|
|
"threshold1Color": "rgba(216, 200, 27, 0.27)",
|
|
"threshold2": null,
|
|
"threshold2Color": "rgba(234, 112, 112, 0.22)"
|
|
},
|
|
"id": 3,
|
|
"isNew": true,
|
|
"legend": {
|
|
"alignAsTable": true,
|
|
"avg": true,
|
|
"current": true,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": true,
|
|
"show": false,
|
|
"sort": "current",
|
|
"sortDesc": true,
|
|
"total": false,
|
|
"values": true
|
|
},
|
|
"lines": true,
|
|
"linewidth": 2,
|
|
"links": [],
|
|
"nullPointMode": "connected",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"seriesOverrides": [],
|
|
"span": 9,
|
|
"stack": false,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "sum(sum by (container_name)(rate(container_cpu_usage_seconds_total{image!=\"\",instance=~\"$node\"}[1m])))",
|
|
"interval": "10s",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "CPU Usage",
|
|
"metric": "container_cpu_usage_seconds_total",
|
|
"refId": "A",
|
|
"step": 10
|
|
},
|
|
{
|
|
"expr": "count(node_cpu{mode=\"system\",instance=~\"$node:9100\"})",
|
|
"interval": "10s",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "CPU Total",
|
|
"metric": "node_cpu",
|
|
"refId": "B",
|
|
"step": 10
|
|
}
|
|
],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "Cluster CPU Usage",
|
|
"tooltip": {
|
|
"msResolution": true,
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "cumulative"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"show": true
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "percentunit",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "short",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
}
|
|
]
|
|
}
|
|
],
|
|
"title": "Row"
|
|
},
|
|
{
|
|
"collapse": false,
|
|
"editable": true,
|
|
"height": "250px",
|
|
"panels": [
|
|
{
|
|
"cacheTimeout": null,
|
|
"colorBackground": false,
|
|
"colorValue": false,
|
|
"colors": [
|
|
"rgba(50, 172, 45, 0.97)",
|
|
"rgba(237, 129, 40, 0.89)",
|
|
"rgba(245, 54, 54, 0.9)"
|
|
],
|
|
"datasource": "${DS_PROMETHEUS}",
|
|
"editable": true,
|
|
"error": false,
|
|
"format": "percent",
|
|
"gauge": {
|
|
"maxValue": 100,
|
|
"minValue": 0,
|
|
"show": true,
|
|
"thresholdLabels": false,
|
|
"thresholdMarkers": true
|
|
},
|
|
"id": 4,
|
|
"interval": null,
|
|
"isNew": true,
|
|
"links": [],
|
|
"mappingType": 1,
|
|
"mappingTypes": [
|
|
{
|
|
"name": "value to text",
|
|
"value": 1
|
|
},
|
|
{
|
|
"name": "range to text",
|
|
"value": 2
|
|
}
|
|
],
|
|
"maxDataPoints": 100,
|
|
"nullPointMode": "connected",
|
|
"nullText": null,
|
|
"postfix": "",
|
|
"postfixFontSize": "50%",
|
|
"prefix": "",
|
|
"prefixFontSize": "50%",
|
|
"rangeMaps": [
|
|
{
|
|
"from": "null",
|
|
"text": "N/A",
|
|
"to": "null"
|
|
}
|
|
],
|
|
"span": 3,
|
|
"sparkline": {
|
|
"fillColor": "rgba(31, 118, 189, 0.18)",
|
|
"full": false,
|
|
"lineColor": "rgb(31, 120, 193)",
|
|
"show": false
|
|
},
|
|
"targets": [
|
|
{
|
|
"expr": "(sum(node_memory_MemTotal{instance=~\"$node:9100\"}) - sum(node_memory_MemFree{instance=~\"$node:9100\"}+node_memory_Buffers{instance=~\"$node:9100\"}+node_memory_Cached{instance=~\"$node:9100\"}) ) / sum(node_memory_MemTotal{instance=~\"$node:9100\"}) * 100",
|
|
"interval": "10s",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "",
|
|
"refId": "A",
|
|
"step": 10
|
|
}
|
|
],
|
|
"thresholds": "65, 90",
|
|
"title": "Cluster memory usage",
|
|
"type": "singlestat",
|
|
"valueFontSize": "80%",
|
|
"valueMaps": [
|
|
{
|
|
"op": "=",
|
|
"text": "N/A",
|
|
"value": "null"
|
|
}
|
|
],
|
|
"valueName": "current"
|
|
},
|
|
{
|
|
"aliasColors": {},
|
|
"bars": false,
|
|
"datasource": "${DS_PROMETHEUS}",
|
|
"decimals": 2,
|
|
"editable": true,
|
|
"error": false,
|
|
"fill": 0,
|
|
"grid": {
|
|
"threshold1": null,
|
|
"threshold1Color": "rgba(216, 200, 27, 0.27)",
|
|
"threshold2": null,
|
|
"threshold2Color": "rgba(234, 112, 112, 0.22)"
|
|
},
|
|
"id": 2,
|
|
"isNew": true,
|
|
"legend": {
|
|
"alignAsTable": true,
|
|
"avg": true,
|
|
"current": true,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": true,
|
|
"show": false,
|
|
"sideWidth": 200,
|
|
"sort": "current",
|
|
"sortDesc": true,
|
|
"total": false,
|
|
"values": true
|
|
},
|
|
"lines": true,
|
|
"linewidth": 2,
|
|
"links": [],
|
|
"nullPointMode": "connected",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"seriesOverrides": [],
|
|
"span": 9,
|
|
"stack": false,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "(sum(node_memory_MemTotal{instance=~\"$node:9100\"}) - sum(node_memory_MemFree{instance=~\"$node:9100\"}+node_memory_Buffers{instance=~\"$node:9100\"}+node_memory_Cached{instance=~\"$node:9100\"}))",
|
|
"interval": "10s",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "Cluster Memory Usage",
|
|
"metric": "node_memory_MemAvailable",
|
|
"refId": "A",
|
|
"step": 10
|
|
},
|
|
{
|
|
"expr": "sum(node_memory_MemTotal{instance=~\"$node:9100\"})",
|
|
"interval": "10s",
|
|
"intervalFactor": 2,
|
|
"legendFormat": "Cluster Memory Total",
|
|
"metric": "node_memory_MemTotal",
|
|
"refId": "B",
|
|
"step": 20
|
|
}
|
|
],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "Cluster Memory Usage",
|
|
"tooltip": {
|
|
"msResolution": false,
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "cumulative"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"show": true
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "bytes",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "short",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
}
|
|
]
|
|
}
|
|
],
|
|
"title": "New row"
|
|
},
|
|
{
|
|
"collapse": false,
|
|
"editable": true,
|
|
"height": "250px",
|
|
"panels": [
|
|
{
|
|
"aliasColors": {},
|
|
"bars": false,
|
|
"datasource": "${DS_PROMETHEUS}",
|
|
"decimals": 2,
|
|
"editable": true,
|
|
"error": false,
|
|
"fill": 0,
|
|
"grid": {
|
|
"threshold1": null,
|
|
"threshold1Color": "rgba(216, 200, 27, 0.27)",
|
|
"threshold2": null,
|
|
"threshold2Color": "rgba(234, 112, 112, 0.22)"
|
|
},
|
|
"id": 9,
|
|
"isNew": true,
|
|
"legend": {
|
|
"alignAsTable": true,
|
|
"avg": true,
|
|
"current": true,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": true,
|
|
"show": true,
|
|
"sideWidth": 200,
|
|
"sort": "current",
|
|
"sortDesc": true,
|
|
"total": true,
|
|
"values": true
|
|
},
|
|
"lines": true,
|
|
"linewidth": 2,
|
|
"links": [],
|
|
"nullPointMode": "connected",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"seriesOverrides": [],
|
|
"span": 12,
|
|
"stack": false,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "sum (rate (node_network_receive_bytes{instance=~\"$node:9100\"}[1m]))",
|
|
"interval": "10s",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "Received Bytes",
|
|
"metric": "network",
|
|
"refId": "A",
|
|
"step": 10
|
|
},
|
|
{
|
|
"expr": "sum (rate (node_network_transmit_bytes{instance=~\"$node:9100\"}[1m]))",
|
|
"interval": "10s",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "Transmitted Bytes",
|
|
"metric": "network",
|
|
"refId": "B",
|
|
"step": 10
|
|
}
|
|
],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "Cluster Network I/O",
|
|
"tooltip": {
|
|
"msResolution": false,
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "cumulative"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"show": true
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "bytes",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "short",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": null,
|
|
"show": true
|
|
}
|
|
]
|
|
}
|
|
],
|
|
"title": "New row"
|
|
}
|
|
],
|
|
"schemaVersion": 12,
|
|
"sharedCrosshair": true,
|
|
"style": "dark",
|
|
"tags": [],
|
|
"templating": {
|
|
"list": [
|
|
{
|
|
"allValue": ".*",
|
|
"current": {},
|
|
"datasource": "${DS_PROMETHEUS}",
|
|
"hide": 0,
|
|
"includeAll": true,
|
|
"label": "Node",
|
|
"multi": false,
|
|
"name": "node",
|
|
"options": [],
|
|
"query": "label_values(kube_node_info, node)",
|
|
"refresh": 1,
|
|
"regex": "",
|
|
"type": "query"
|
|
}
|
|
]
|
|
},
|
|
"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": "browser",
|
|
"title": "Kubernetes Cluster Monitoring",
|
|
"version": 0
|
|
},
|
|
"inputs": [
|
|
{
|
|
"name": "DS_PROMETHEUS",
|
|
"pluginId": "prometheus",
|
|
"type": "datasource",
|
|
"value": "prometheus"
|
|
}
|
|
],
|
|
"overwrite": true
|
|
}
|