Files
yucca/o11y/dashboards/yucca-spice-ceph-health.json
T

1271 lines
29 KiB
JSON

{
"annotations": {
"list": [
{
"builtIn": 1,
"datasource": {
"uid": "-- Grafana --"
},
"enable": true,
"hide": true,
"iconColor": "rgba(0, 211, 255, 1)",
"name": "Annotations & Alerts",
"type": "dashboard"
}
]
},
"editable": true,
"fiscalYearStartMonth": 0,
"graphTooltip": 1,
"links": [],
"panels": [
{
"collapsed": false,
"gridPos": {
"h": 1,
"w": 24,
"x": 0,
"y": 0
},
"id": 1,
"panels": [],
"title": "Cluster state",
"type": "row"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "thresholds"
},
"mappings": [
{
"options": {
"0": {
"color": "green",
"index": 0,
"text": "OK"
},
"1": {
"color": "yellow",
"index": 1,
"text": "WARN"
},
"2": {
"color": "red",
"index": 2,
"text": "ERR"
}
},
"type": "value"
}
],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
},
{
"color": "yellow",
"value": 1
},
{
"color": "red",
"value": 2
}
]
},
"unit": "short"
},
"overrides": []
},
"gridPos": {
"h": 5,
"w": 4,
"x": 0,
"y": 1
},
"id": 2,
"options": {
"colorMode": "value",
"graphMode": "area",
"justifyMode": "auto",
"orientation": "auto",
"reduceOptions": {
"calcs": [
"lastNotNull"
],
"fields": "",
"values": false
},
"textMode": "auto",
"wideLayout": true
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "ceph_health_status{cluster=~\"$cluster\"}"
}
],
"title": "Health",
"type": "stat"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "thresholds"
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "red",
"value": 0
},
{
"color": "yellow",
"value": 3
},
{
"color": "green",
"value": 5
}
]
},
"unit": "short"
},
"overrides": []
},
"gridPos": {
"h": 5,
"w": 4,
"x": 4,
"y": 1
},
"id": 3,
"options": {
"colorMode": "value",
"graphMode": "area",
"justifyMode": "auto",
"orientation": "auto",
"reduceOptions": {
"calcs": [
"lastNotNull"
],
"fields": "",
"values": false
},
"textMode": "auto",
"wideLayout": true
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "sum(ceph_mon_quorum_status{cluster=~\"$cluster\"})"
}
],
"title": "Mons in quorum",
"type": "stat"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "thresholds"
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "red",
"value": 0
},
{
"color": "green",
"value": 500
}
]
},
"unit": "short"
},
"overrides": []
},
"gridPos": {
"h": 5,
"w": 4,
"x": 8,
"y": 1
},
"id": 4,
"options": {
"colorMode": "value",
"graphMode": "area",
"justifyMode": "auto",
"orientation": "auto",
"reduceOptions": {
"calcs": [
"lastNotNull"
],
"fields": "",
"values": false
},
"textMode": "auto",
"wideLayout": true
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "sum(ceph_osd_up{cluster=~\"$cluster\"})"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "B",
"expr": "count(ceph_osd_up{cluster=~\"$cluster\"})"
}
],
"title": "OSDs up / total",
"type": "stat"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "thresholds"
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
}
]
},
"unit": "short"
},
"overrides": []
},
"gridPos": {
"h": 5,
"w": 4,
"x": 12,
"y": 1
},
"id": 5,
"options": {
"colorMode": "value",
"graphMode": "area",
"justifyMode": "auto",
"orientation": "auto",
"reduceOptions": {
"calcs": [
"lastNotNull"
],
"fields": "",
"values": false
},
"textMode": "auto",
"wideLayout": true
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "sum(ceph_osd_in{cluster=~\"$cluster\"})"
}
],
"title": "OSDs in",
"type": "stat"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "thresholds"
},
"mappings": [],
"max": 1,
"min": 0,
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
},
{
"color": "yellow",
"value": 0.7
},
{
"color": "orange",
"value": 0.8
},
{
"color": "red",
"value": 0.9
}
]
},
"unit": "percentunit"
},
"overrides": []
},
"gridPos": {
"h": 5,
"w": 4,
"x": 16,
"y": 1
},
"id": 6,
"options": {
"minVizHeight": 75,
"minVizWidth": 75,
"orientation": "auto",
"reduceOptions": {
"calcs": [
"lastNotNull"
],
"fields": "",
"values": false
},
"showThresholdLabels": false,
"showThresholdMarkers": true,
"sizing": "auto"
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "sum(ceph_cluster_total_used_bytes{cluster=~\"$cluster\"}) / sum(ceph_cluster_total_bytes{cluster=~\"$cluster\"})"
}
],
"title": "Raw used",
"type": "gauge",
"description": "Raw includes EC parity + block.db; use pool % for planning"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "thresholds"
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
}
]
},
"unit": "short"
},
"overrides": []
},
"gridPos": {
"h": 5,
"w": 4,
"x": 20,
"y": 1
},
"id": 7,
"options": {
"colorMode": "value",
"graphMode": "area",
"justifyMode": "auto",
"orientation": "auto",
"reduceOptions": {
"calcs": [
"lastNotNull"
],
"fields": "",
"values": false
},
"textMode": "auto",
"wideLayout": true
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "sum(ceph_pool_objects{cluster=~\"$cluster\"})"
}
],
"title": "Objects",
"type": "stat"
},
{
"collapsed": false,
"gridPos": {
"h": 1,
"w": 24,
"x": 0,
"y": 6
},
"id": 8,
"panels": [],
"title": "Placement groups",
"type": "row"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "palette-classic"
},
"custom": {
"axisCenteredZero": false,
"axisPlacement": "auto",
"drawStyle": "line",
"fillOpacity": 0,
"gradientMode": "none",
"lineInterpolation": "linear",
"lineWidth": 2,
"pointSize": 5,
"scaleDistribution": {
"type": "linear"
},
"showPoints": "never",
"spanNulls": true,
"stacking": {
"group": "A",
"mode": "none"
},
"thresholdsStyle": {
"mode": "off"
}
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
}
]
},
"unit": "short"
},
"overrides": []
},
"gridPos": {
"h": 8,
"w": 12,
"x": 0,
"y": 7
},
"id": 9,
"options": {
"legend": {
"calcs": [
"lastNotNull"
],
"displayMode": "table",
"placement": "bottom",
"showLegend": true,
"sortBy": "Last *",
"sortDesc": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "sum(ceph_pg_degraded{cluster=~\"$cluster\"})",
"legendFormat": "degraded"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "B",
"expr": "sum(ceph_pg_undersized{cluster=~\"$cluster\"})",
"legendFormat": "undersized"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "C",
"expr": "sum(ceph_pg_peering{cluster=~\"$cluster\"})",
"legendFormat": "peering"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "D",
"expr": "sum(ceph_pg_backfilling{cluster=~\"$cluster\"})",
"legendFormat": "backfilling"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "E",
"expr": "sum(ceph_pg_recovering{cluster=~\"$cluster\"})",
"legendFormat": "recovering"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "F",
"expr": "sum(ceph_pg_inconsistent{cluster=~\"$cluster\"})",
"legendFormat": "inconsistent"
}
],
"title": "PG states (unhealthy)",
"type": "timeseries",
"description": "All zero on a healthy cluster"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "palette-classic"
},
"custom": {
"axisCenteredZero": false,
"axisPlacement": "auto",
"drawStyle": "line",
"fillOpacity": 0,
"gradientMode": "none",
"lineInterpolation": "linear",
"lineWidth": 2,
"pointSize": 5,
"scaleDistribution": {
"type": "linear"
},
"showPoints": "never",
"spanNulls": true,
"stacking": {
"group": "A",
"mode": "none"
},
"thresholdsStyle": {
"mode": "off"
}
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
}
]
},
"unit": "short"
},
"overrides": []
},
"gridPos": {
"h": 8,
"w": 12,
"x": 12,
"y": 7
},
"id": 10,
"options": {
"legend": {
"calcs": [
"lastNotNull"
],
"displayMode": "table",
"placement": "bottom",
"showLegend": true,
"sortBy": "Last *",
"sortDesc": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "sum(ceph_pg_active{cluster=~\"$cluster\"})",
"legendFormat": "active"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "B",
"expr": "sum(ceph_pg_clean{cluster=~\"$cluster\"})",
"legendFormat": "clean"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "C",
"expr": "sum(ceph_pg_total{cluster=~\"$cluster\"})",
"legendFormat": "total"
}
],
"title": "PGs active/clean vs total",
"type": "timeseries"
},
{
"collapsed": false,
"gridPos": {
"h": 1,
"w": 24,
"x": 0,
"y": 15
},
"id": 11,
"panels": [],
"title": "Client + recovery IO",
"type": "row"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "palette-classic"
},
"custom": {
"axisCenteredZero": false,
"axisPlacement": "auto",
"drawStyle": "line",
"fillOpacity": 0,
"gradientMode": "none",
"lineInterpolation": "linear",
"lineWidth": 2,
"pointSize": 5,
"scaleDistribution": {
"type": "linear"
},
"showPoints": "never",
"spanNulls": true,
"stacking": {
"group": "A",
"mode": "none"
},
"thresholdsStyle": {
"mode": "off"
}
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
}
]
},
"unit": "Bps"
},
"overrides": []
},
"gridPos": {
"h": 8,
"w": 12,
"x": 0,
"y": 16
},
"id": 12,
"options": {
"legend": {
"calcs": [
"lastNotNull"
],
"displayMode": "table",
"placement": "bottom",
"showLegend": true,
"sortBy": "Last *",
"sortDesc": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "sum(rate(ceph_pool_rd_bytes{cluster=~\"$cluster\"}[$__rate_interval]))",
"legendFormat": "read"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "B",
"expr": "sum(rate(ceph_pool_wr_bytes{cluster=~\"$cluster\"}[$__rate_interval]))",
"legendFormat": "write"
}
],
"title": "Client throughput",
"type": "timeseries"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "palette-classic"
},
"custom": {
"axisCenteredZero": false,
"axisPlacement": "auto",
"drawStyle": "line",
"fillOpacity": 0,
"gradientMode": "none",
"lineInterpolation": "linear",
"lineWidth": 2,
"pointSize": 5,
"scaleDistribution": {
"type": "linear"
},
"showPoints": "never",
"spanNulls": true,
"stacking": {
"group": "A",
"mode": "none"
},
"thresholdsStyle": {
"mode": "off"
}
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
}
]
},
"unit": "iops"
},
"overrides": []
},
"gridPos": {
"h": 8,
"w": 12,
"x": 12,
"y": 16
},
"id": 13,
"options": {
"legend": {
"calcs": [
"lastNotNull"
],
"displayMode": "table",
"placement": "bottom",
"showLegend": true,
"sortBy": "Last *",
"sortDesc": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "sum(rate(ceph_pool_rd{cluster=~\"$cluster\"}[$__rate_interval]))",
"legendFormat": "read"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "B",
"expr": "sum(rate(ceph_pool_wr{cluster=~\"$cluster\"}[$__rate_interval]))",
"legendFormat": "write"
}
],
"title": "Client IOPS",
"type": "timeseries"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "palette-classic"
},
"custom": {
"axisCenteredZero": false,
"axisPlacement": "auto",
"drawStyle": "line",
"fillOpacity": 0,
"gradientMode": "none",
"lineInterpolation": "linear",
"lineWidth": 2,
"pointSize": 5,
"scaleDistribution": {
"type": "linear"
},
"showPoints": "never",
"spanNulls": true,
"stacking": {
"group": "A",
"mode": "none"
},
"thresholdsStyle": {
"mode": "off"
}
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
}
]
},
"unit": "ops"
},
"overrides": []
},
"gridPos": {
"h": 8,
"w": 12,
"x": 0,
"y": 24
},
"id": 14,
"options": {
"legend": {
"calcs": [
"lastNotNull"
],
"displayMode": "table",
"placement": "bottom",
"showLegend": true,
"sortBy": "Last *",
"sortDesc": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "sum(rate(ceph_osd_recovery_ops{cluster=~\"$cluster\"}[$__rate_interval]))",
"legendFormat": "recovery"
}
],
"title": "Recovery ops/s",
"type": "timeseries",
"description": "Non-zero means rebalance/recovery is running"
},
{
"collapsed": false,
"gridPos": {
"h": 1,
"w": 24,
"x": 0,
"y": 32
},
"id": 15,
"panels": [],
"title": "OSD latency + fullness",
"type": "row"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "palette-classic"
},
"custom": {
"axisCenteredZero": false,
"axisPlacement": "auto",
"drawStyle": "line",
"fillOpacity": 0,
"gradientMode": "none",
"lineInterpolation": "linear",
"lineWidth": 2,
"pointSize": 5,
"scaleDistribution": {
"type": "linear"
},
"showPoints": "never",
"spanNulls": true,
"stacking": {
"group": "A",
"mode": "none"
},
"thresholdsStyle": {
"mode": "off"
}
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
}
]
},
"unit": "ms"
},
"overrides": []
},
"gridPos": {
"h": 8,
"w": 12,
"x": 0,
"y": 33
},
"id": 16,
"options": {
"legend": {
"calcs": [
"lastNotNull"
],
"displayMode": "table",
"placement": "bottom",
"showLegend": true,
"sortBy": "Last *",
"sortDesc": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "avg(ceph_osd_apply_latency_ms{cluster=~\"$cluster\"})",
"legendFormat": "apply (avg)"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "B",
"expr": "avg(ceph_osd_commit_latency_ms{cluster=~\"$cluster\"})",
"legendFormat": "commit (avg)"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "C",
"expr": "max(ceph_osd_apply_latency_ms{cluster=~\"$cluster\"})",
"legendFormat": "apply (worst OSD)"
},
{
"datasource": {
"uid": "$datasource"
},
"refId": "D",
"expr": "max(ceph_osd_commit_latency_ms{cluster=~\"$cluster\"})",
"legendFormat": "commit (worst OSD)"
}
],
"title": "OSD apply/commit latency",
"type": "timeseries"
},
{
"datasource": {
"uid": "$datasource"
},
"fieldConfig": {
"defaults": {
"color": {
"mode": "palette-classic"
},
"custom": {
"axisCenteredZero": false,
"axisPlacement": "auto",
"drawStyle": "line",
"fillOpacity": 0,
"gradientMode": "none",
"lineInterpolation": "linear",
"lineWidth": 2,
"pointSize": 5,
"scaleDistribution": {
"type": "linear"
},
"showPoints": "never",
"spanNulls": true,
"stacking": {
"group": "A",
"mode": "none"
},
"thresholdsStyle": {
"mode": "off"
}
},
"mappings": [],
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": 0
}
]
},
"unit": "percentunit",
"min": 0,
"max": 1
},
"overrides": []
},
"gridPos": {
"h": 8,
"w": 12,
"x": 12,
"y": 33
},
"id": 17,
"options": {
"legend": {
"calcs": [
"lastNotNull"
],
"displayMode": "table",
"placement": "bottom",
"showLegend": true,
"sortBy": "Last *",
"sortDesc": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
},
"targets": [
{
"datasource": {
"uid": "$datasource"
},
"refId": "A",
"expr": "topk(10, ceph_osd_stat_bytes_used{cluster=~\"$cluster\"} / ceph_osd_stat_bytes{cluster=~\"$cluster\"})",
"legendFormat": "{{ceph_daemon}}"
}
],
"title": "Fullest OSDs (top 10)",
"type": "timeseries"
}
],
"preload": false,
"refresh": "30s",
"schemaVersion": 39,
"tags": [
"spice",
"ceph",
"prod"
],
"templating": {
"list": [
{
"current": {
"text": "VictoriaMetrics Fleet",
"value": "VictoriaMetricsFleet"
},
"label": "Data source",
"name": "datasource",
"options": [],
"query": "prometheus",
"refresh": 1,
"regex": "/^VictoriaMetrics Fleet$/",
"type": "datasource"
},
{
"allValue": ".*",
"current": {
"selected": true,
"text": [
"All"
],
"value": [
"$__all"
]
},
"datasource": {
"uid": "$datasource"
},
"includeAll": true,
"label": "Env",
"multi": true,
"name": "env",
"options": [],
"query": "label_values(up{project=\"yucca\"}, env)",
"refresh": 1,
"regex": "",
"sort": 1,
"type": "query"
},
{
"allValue": ".*",
"current": {
"selected": true,
"text": [
"All"
],
"value": [
"$__all"
]
},
"datasource": {
"uid": "$datasource"
},
"includeAll": true,
"label": "Provider",
"multi": true,
"name": "provider",
"options": [],
"query": "label_values(up{project=\"yucca\", env=~\"$env\"}, provider)",
"refresh": 1,
"regex": "",
"sort": 1,
"type": "query"
},
{
"allValue": ".*",
"current": {
"selected": true,
"text": [
"All"
],
"value": [
"$__all"
]
},
"datasource": {
"uid": "$datasource"
},
"includeAll": true,
"label": "Region",
"multi": true,
"name": "region",
"options": [],
"query": "label_values(up{project=\"yucca\", env=~\"$env\", provider=~\"$provider\"}, region)",
"refresh": 1,
"regex": "",
"sort": 1,
"type": "query"
},
{
"current": {},
"datasource": {
"uid": "$datasource"
},
"includeAll": false,
"label": "cluster",
"name": "cluster",
"options": [],
"query": "label_values(ceph_health_status{env=~\"$env\", provider=~\"$provider\", region=~\"$region\"}, cluster)",
"refresh": 1,
"regex": "",
"sort": 1,
"type": "query"
}
]
},
"time": {
"from": "now-6h",
"to": "now"
},
"timepicker": {},
"timezone": "",
"title": "Ceph / Health",
"uid": "yucca-spice-ceph-health",
"version": 1
}