e35d9082d5
This supports grafana dashboards and metrics scraping from the ceph mgr prometheus module. Change-Id: I8b2f132a4997d205119f7afe2a1ab6b2ae4c0134
696 lines
19 KiB
JSON
696 lines
19 KiB
JSON
{
|
|
"__inputs": [ ],
|
|
"__requires": [
|
|
{
|
|
"id": "grafana",
|
|
"name": "Grafana",
|
|
"type": "grafana",
|
|
"version": "5.0.0"
|
|
},
|
|
{
|
|
"id": "graph",
|
|
"name": "Graph",
|
|
"type": "panel",
|
|
"version": "5.0.0"
|
|
}
|
|
],
|
|
"annotations": {
|
|
"list": [
|
|
{
|
|
"builtIn": 1,
|
|
"datasource": "-- Grafana --",
|
|
"enable": true,
|
|
"hide": true,
|
|
"iconColor": "rgba(0, 211, 255, 1)",
|
|
"name": "Annotations & Alerts",
|
|
"showIn": 0,
|
|
"tags": [ ],
|
|
"type": "dashboard"
|
|
}
|
|
]
|
|
},
|
|
"description": "",
|
|
"editable": false,
|
|
"gnetId": null,
|
|
"graphTooltip": 0,
|
|
"hideControls": false,
|
|
"id": null,
|
|
"links": [ ],
|
|
"panels": [
|
|
{
|
|
"collapse": false,
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 0
|
|
},
|
|
"id": 2,
|
|
"panels": [ ],
|
|
"repeat": null,
|
|
"repeatIteration": null,
|
|
"repeatRowId": null,
|
|
"showTitle": true,
|
|
"title": "RGW Overview - All Gateways",
|
|
"titleSize": "h6",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"aliasColors": { },
|
|
"bars": false,
|
|
"dashLength": 10,
|
|
"dashes": false,
|
|
"datasource": "${prometheusds}",
|
|
"description": "",
|
|
"fill": 1,
|
|
"fillGradient": 0,
|
|
"gridPos": {
|
|
"h": 7,
|
|
"w": 8,
|
|
"x": 0,
|
|
"y": 1
|
|
},
|
|
"id": 3,
|
|
"legend": {
|
|
"alignAsTable": false,
|
|
"avg": false,
|
|
"current": false,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": false,
|
|
"show": true,
|
|
"sideWidth": null,
|
|
"total": false,
|
|
"values": false
|
|
},
|
|
"lines": true,
|
|
"linewidth": 1,
|
|
"links": [ ],
|
|
"nullPointMode": "null as zero",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"repeat": null,
|
|
"seriesOverrides": [ ],
|
|
"spaceLength": 10,
|
|
"stack": false,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "label_replace(\n rate(ceph_rgw_get_initial_lat_sum{job=~\"$job\"}[$__rate_interval]) /\n rate(ceph_rgw_get_initial_lat_count{job=~\"$job\"}[$__rate_interval]) *\n on (instance_id) group_left (ceph_daemon) ceph_rgw_metadata{job=~\"$job\"},\n \"rgw_host\", \"$1\", \"ceph_daemon\", \"rgw.(.*)\"\n)\n",
|
|
"format": "time_series",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "GET {{rgw_host}}",
|
|
"refId": "A"
|
|
},
|
|
{
|
|
"expr": "label_replace(\n rate(ceph_rgw_put_initial_lat_sum{job=~\"$job\"}[$__rate_interval]) /\n rate(ceph_rgw_put_initial_lat_count{job=~\"$job\"}[$__rate_interval]) *\n on (instance_id) group_left (ceph_daemon) ceph_rgw_metadata{job=~\"$job\"},\n \"rgw_host\", \"$1\", \"ceph_daemon\", \"rgw.(.*)\"\n)\n",
|
|
"format": "time_series",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "PUT {{rgw_host}}",
|
|
"refId": "B"
|
|
}
|
|
],
|
|
"thresholds": [ ],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "Average GET/PUT Latencies by RGW Instance",
|
|
"tooltip": {
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "individual"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"buckets": null,
|
|
"mode": "time",
|
|
"name": null,
|
|
"show": true,
|
|
"values": [ ]
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "s",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "short",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"aliasColors": { },
|
|
"bars": false,
|
|
"dashLength": 10,
|
|
"dashes": false,
|
|
"datasource": "${prometheusds}",
|
|
"description": "",
|
|
"fill": 1,
|
|
"fillGradient": 0,
|
|
"gridPos": {
|
|
"h": 7,
|
|
"w": 7,
|
|
"x": 8,
|
|
"y": 1
|
|
},
|
|
"id": 4,
|
|
"legend": {
|
|
"alignAsTable": false,
|
|
"avg": false,
|
|
"current": false,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": false,
|
|
"show": true,
|
|
"sideWidth": null,
|
|
"total": false,
|
|
"values": false
|
|
},
|
|
"lines": true,
|
|
"linewidth": 1,
|
|
"links": [ ],
|
|
"nullPointMode": "null as zero",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"repeat": null,
|
|
"seriesOverrides": [ ],
|
|
"spaceLength": 10,
|
|
"stack": false,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "sum by (rgw_host) (\n label_replace(\n rate(ceph_rgw_req{job=~\"$job\"}[$__rate_interval]) *\n on (instance_id) group_left (ceph_daemon) ceph_rgw_metadata{job=~\"$job\"},\n \"rgw_host\", \"$1\", \"ceph_daemon\", \"rgw.(.*)\"\n )\n)\n",
|
|
"format": "time_series",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "{{rgw_host}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"thresholds": [ ],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "Total Requests/sec by RGW Instance",
|
|
"tooltip": {
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "individual"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"buckets": null,
|
|
"mode": "time",
|
|
"name": null,
|
|
"show": true,
|
|
"values": [ ]
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "none",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "short",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"aliasColors": { },
|
|
"bars": false,
|
|
"dashLength": 10,
|
|
"dashes": false,
|
|
"datasource": "${prometheusds}",
|
|
"description": "Latencies are shown stacked, without a yaxis to provide a visual indication of GET latency imbalance across RGW hosts",
|
|
"fill": 1,
|
|
"fillGradient": 0,
|
|
"gridPos": {
|
|
"h": 7,
|
|
"w": 6,
|
|
"x": 15,
|
|
"y": 1
|
|
},
|
|
"id": 5,
|
|
"legend": {
|
|
"alignAsTable": false,
|
|
"avg": false,
|
|
"current": false,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": false,
|
|
"show": true,
|
|
"sideWidth": null,
|
|
"total": false,
|
|
"values": false
|
|
},
|
|
"lines": true,
|
|
"linewidth": 1,
|
|
"links": [ ],
|
|
"nullPointMode": "null as zero",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"repeat": null,
|
|
"seriesOverrides": [ ],
|
|
"spaceLength": 10,
|
|
"stack": false,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "label_replace(\n rate(ceph_rgw_get_initial_lat_sum{job=~\"$job\"}[$__rate_interval]) /\n rate(ceph_rgw_get_initial_lat_count{job=~\"$job\"}[$__rate_interval]) *\n on (instance_id) group_left (ceph_daemon) ceph_rgw_metadata{job=~\"$job\"},\n \"rgw_host\", \"$1\", \"ceph_daemon\", \"rgw.(.*)\"\n)\n",
|
|
"format": "time_series",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "{{rgw_host}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"thresholds": [ ],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "GET Latencies by RGW Instance",
|
|
"tooltip": {
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "individual"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"buckets": null,
|
|
"mode": "time",
|
|
"name": null,
|
|
"show": true,
|
|
"values": [ ]
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "s",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "short",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"aliasColors": { },
|
|
"bars": false,
|
|
"dashLength": 10,
|
|
"dashes": false,
|
|
"datasource": "${prometheusds}",
|
|
"description": "Total bytes transferred in/out of all radosgw instances within the cluster",
|
|
"fill": 1,
|
|
"fillGradient": 0,
|
|
"gridPos": {
|
|
"h": 6,
|
|
"w": 8,
|
|
"x": 0,
|
|
"y": 8
|
|
},
|
|
"id": 6,
|
|
"legend": {
|
|
"alignAsTable": false,
|
|
"avg": false,
|
|
"current": false,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": false,
|
|
"show": true,
|
|
"sideWidth": null,
|
|
"total": false,
|
|
"values": false
|
|
},
|
|
"lines": true,
|
|
"linewidth": 1,
|
|
"links": [ ],
|
|
"nullPointMode": "null as zero",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"repeat": null,
|
|
"seriesOverrides": [ ],
|
|
"spaceLength": 10,
|
|
"stack": false,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "sum(rate(ceph_rgw_get_b{job=~\"$job\"}[$__rate_interval]))",
|
|
"format": "time_series",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "GETs",
|
|
"refId": "A"
|
|
},
|
|
{
|
|
"expr": "sum(rate(ceph_rgw_put_b{job=~\"$job\"}[$__rate_interval]))",
|
|
"format": "time_series",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "PUTs",
|
|
"refId": "B"
|
|
}
|
|
],
|
|
"thresholds": [ ],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "Bandwidth Consumed by Type",
|
|
"tooltip": {
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "individual"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"buckets": null,
|
|
"mode": "time",
|
|
"name": null,
|
|
"show": true,
|
|
"values": [ ]
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "bytes",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "short",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"aliasColors": { },
|
|
"bars": false,
|
|
"dashLength": 10,
|
|
"dashes": false,
|
|
"datasource": "${prometheusds}",
|
|
"description": "Total bytes transferred in/out through get/put operations, by radosgw instance",
|
|
"fill": 1,
|
|
"fillGradient": 0,
|
|
"gridPos": {
|
|
"h": 6,
|
|
"w": 7,
|
|
"x": 8,
|
|
"y": 8
|
|
},
|
|
"id": 7,
|
|
"legend": {
|
|
"alignAsTable": false,
|
|
"avg": false,
|
|
"current": false,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": false,
|
|
"show": true,
|
|
"sideWidth": null,
|
|
"total": false,
|
|
"values": false
|
|
},
|
|
"lines": true,
|
|
"linewidth": 1,
|
|
"links": [ ],
|
|
"nullPointMode": "null as zero",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"repeat": null,
|
|
"seriesOverrides": [ ],
|
|
"spaceLength": 10,
|
|
"stack": false,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "label_replace(sum by (instance_id) (\n rate(ceph_rgw_get_b{job=~\"$job\"}[$__rate_interval]) +\n rate(ceph_rgw_put_b{job=~\"$job\"}[$__rate_interval])) *\n on (instance_id) group_left (ceph_daemon) ceph_rgw_metadata{job=~\"$job\"},\n \"rgw_host\", \"$1\", \"ceph_daemon\", \"rgw.(.*)\"\n)\n",
|
|
"format": "time_series",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "{{rgw_host}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"thresholds": [ ],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "Bandwidth by RGW Instance",
|
|
"tooltip": {
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "individual"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"buckets": null,
|
|
"mode": "time",
|
|
"name": null,
|
|
"show": true,
|
|
"values": [ ]
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "bytes",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "short",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"aliasColors": { },
|
|
"bars": false,
|
|
"dashLength": 10,
|
|
"dashes": false,
|
|
"datasource": "${prometheusds}",
|
|
"description": "Latencies are shown stacked, without a yaxis to provide a visual indication of PUT latency imbalance across RGW hosts",
|
|
"fill": 1,
|
|
"fillGradient": 0,
|
|
"gridPos": {
|
|
"h": 6,
|
|
"w": 6,
|
|
"x": 15,
|
|
"y": 8
|
|
},
|
|
"id": 8,
|
|
"legend": {
|
|
"alignAsTable": false,
|
|
"avg": false,
|
|
"current": false,
|
|
"max": false,
|
|
"min": false,
|
|
"rightSide": false,
|
|
"show": true,
|
|
"sideWidth": null,
|
|
"total": false,
|
|
"values": false
|
|
},
|
|
"lines": true,
|
|
"linewidth": 1,
|
|
"links": [ ],
|
|
"nullPointMode": "null as zero",
|
|
"percentage": false,
|
|
"pointradius": 5,
|
|
"points": false,
|
|
"renderer": "flot",
|
|
"repeat": null,
|
|
"seriesOverrides": [ ],
|
|
"spaceLength": 10,
|
|
"stack": false,
|
|
"steppedLine": false,
|
|
"targets": [
|
|
{
|
|
"expr": "label_replace(\n rate(ceph_rgw_put_initial_lat_sum{job=~\"$job\"}[$__rate_interval]) /\n rate(ceph_rgw_put_initial_lat_count{job=~\"$job\"}[$__rate_interval]) *\n on (instance_id) group_left (ceph_daemon) ceph_rgw_metadata{job=~\"$job\"},\n \"rgw_host\", \"$1\", \"ceph_daemon\", \"rgw.(.*)\"\n)\n",
|
|
"format": "time_series",
|
|
"intervalFactor": 1,
|
|
"legendFormat": "{{rgw_host}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"thresholds": [ ],
|
|
"timeFrom": null,
|
|
"timeShift": null,
|
|
"title": "PUT Latencies by RGW Instance",
|
|
"tooltip": {
|
|
"shared": true,
|
|
"sort": 0,
|
|
"value_type": "individual"
|
|
},
|
|
"type": "graph",
|
|
"xaxis": {
|
|
"buckets": null,
|
|
"mode": "time",
|
|
"name": null,
|
|
"show": true,
|
|
"values": [ ]
|
|
},
|
|
"yaxes": [
|
|
{
|
|
"format": "s",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
},
|
|
{
|
|
"format": "short",
|
|
"label": null,
|
|
"logBase": 1,
|
|
"max": null,
|
|
"min": 0,
|
|
"show": true
|
|
}
|
|
]
|
|
}
|
|
],
|
|
"refresh": "30s",
|
|
"rows": [ ],
|
|
"schemaVersion": 16,
|
|
"style": "dark",
|
|
"tags": [
|
|
"ceph-mixin",
|
|
"overview"
|
|
],
|
|
"templating": {
|
|
"list": [
|
|
{
|
|
"allValue": ".+",
|
|
"current": { },
|
|
"datasource": "${prometheusds}",
|
|
"hide": 2,
|
|
"includeAll": true,
|
|
"label": "cluster",
|
|
"multi": true,
|
|
"name": "cluster",
|
|
"options": [ ],
|
|
"query": "label_values(ceph_osd_metadata, cluster)",
|
|
"refresh": 1,
|
|
"regex": "(.*)",
|
|
"sort": 1,
|
|
"tagValuesQuery": "",
|
|
"tags": [ ],
|
|
"tagsQuery": "",
|
|
"type": "query",
|
|
"useTags": false
|
|
},
|
|
{
|
|
"allValue": ".+",
|
|
"current": { },
|
|
"datasource": "${prometheusds}",
|
|
"hide": 0,
|
|
"includeAll": true,
|
|
"label": "job",
|
|
"multi": true,
|
|
"name": "job",
|
|
"options": [ ],
|
|
"query": "label_values(ceph_osd_metadata{}, job)",
|
|
"refresh": 1,
|
|
"regex": "(.*)",
|
|
"sort": 1,
|
|
"tagValuesQuery": "",
|
|
"tags": [ ],
|
|
"tagsQuery": "",
|
|
"type": "query",
|
|
"useTags": false
|
|
},
|
|
{
|
|
"allValue": null,
|
|
"current": { },
|
|
"datasource": "${prometheusds}",
|
|
"hide": 0,
|
|
"includeAll": true,
|
|
"label": "",
|
|
"multi": false,
|
|
"name": "rgw_servers",
|
|
"options": [ ],
|
|
"query": "label_values(ceph_rgw_metadata{job=~\"$job\"}, ceph_daemon)",
|
|
"refresh": 1,
|
|
"regex": "RGW Server",
|
|
"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": "",
|
|
"title": "RGW Overview",
|
|
"uid": "WAkugZpiz",
|
|
"version": 0
|
|
}
|