mirror of
https://github.com/immich-app/yucca.git
synced 2026-09-30 13:33:00 +08:00
1328 lines
34 KiB
JSON
1328 lines
34 KiB
JSON
{
|
|
"uid": "yucca-spice-blockdb-bluefs",
|
|
"title": "Ceph / block.db + BlueFS",
|
|
"description": "Per-OSD block.db utilization, spillover, and the derived sizing constant for the spice cluster. Filtered to device_class=hdd throughout. Provisioned by ansible (ceph_deploy); edit the file in the repo, not in Grafana.",
|
|
"tags": [
|
|
"spice",
|
|
"ceph",
|
|
"bluestore",
|
|
"bluefs",
|
|
"prod"
|
|
],
|
|
"timezone": "",
|
|
"editable": true,
|
|
"graphTooltip": 1,
|
|
"schemaVersion": 39,
|
|
"version": 1,
|
|
"refresh": "1m",
|
|
"time": {
|
|
"from": "now-7d",
|
|
"to": "now"
|
|
},
|
|
"timepicker": {},
|
|
"annotations": {
|
|
"list": []
|
|
},
|
|
"links": [],
|
|
"templating": {
|
|
"list": [
|
|
{
|
|
"name": "datasource",
|
|
"type": "datasource",
|
|
"query": "prometheus",
|
|
"label": "Data Source",
|
|
"refresh": 1,
|
|
"hide": 0,
|
|
"current": {
|
|
"text": "VictoriaMetrics Fleet",
|
|
"value": "VictoriaMetricsFleet"
|
|
},
|
|
"options": [],
|
|
"regex": "/^VictoriaMetrics Fleet$/",
|
|
"skipUrlSync": false
|
|
},
|
|
{
|
|
"allValue": ".*",
|
|
"current": {
|
|
"selected": true,
|
|
"text": [
|
|
"All"
|
|
],
|
|
"value": [
|
|
"$__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"
|
|
},
|
|
{
|
|
"name": "cluster",
|
|
"type": "query",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"query": "label_values(ceph_health_status{env=~\"$env\", provider=~\"$provider\", region=~\"$region\"}, cluster)",
|
|
"label": "Cluster",
|
|
"refresh": 1,
|
|
"hide": 0,
|
|
"includeAll": false,
|
|
"multi": false,
|
|
"allValue": null,
|
|
"current": {},
|
|
"options": [],
|
|
"regex": "",
|
|
"sort": 1,
|
|
"skipUrlSync": false
|
|
}
|
|
]
|
|
},
|
|
"panels": [
|
|
{
|
|
"id": 1,
|
|
"type": "text",
|
|
"title": "How to read this",
|
|
"gridPos": {
|
|
"h": 4,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 0
|
|
},
|
|
"options": {
|
|
"mode": "markdown",
|
|
"content": "Panels cover HDD OSDs only. SSD OSDs are colocated, so their `ceph_bluefs_db_total_bytes` is the whole data device rather than a block.db LV, and including them makes utilization look idle.\n\nWatch spillover first: an OSD with `slow_used_bytes > 0` has filled its block.db and is writing RocksDB onto the spinner. Utilization is the early warning. The volume selector exposes 122 GiB of each 128 GiB LV, so real pressure runs slightly above the percentages shown here."
|
|
}
|
|
},
|
|
{
|
|
"id": 2,
|
|
"type": "row",
|
|
"title": "Spillover tripwire",
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 4
|
|
},
|
|
"panels": []
|
|
},
|
|
{
|
|
"id": 3,
|
|
"type": "stat",
|
|
"title": "OSDs spilling to HDD",
|
|
"description": "HDD OSDs with slow_used_bytes > 0. Anything above zero means block.db is exhausted on that OSD and RocksDB has spilled onto the data device.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 6,
|
|
"x": 0,
|
|
"y": 5
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "count((ceph_bluefs_slow_used_bytes{cluster=~\"$cluster\"} > 0) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"}) or vector(0)",
|
|
"refId": "A",
|
|
"legendFormat": ""
|
|
}
|
|
],
|
|
"options": {
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
],
|
|
"fields": "",
|
|
"values": false
|
|
},
|
|
"orientation": "auto",
|
|
"textMode": "auto",
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"justifyMode": "auto"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "none",
|
|
"decimals": 0,
|
|
"mappings": [],
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
},
|
|
{
|
|
"color": "red",
|
|
"value": 1
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
}
|
|
},
|
|
{
|
|
"id": 4,
|
|
"type": "stat",
|
|
"title": "Worst spillover",
|
|
"description": "Largest slow_used_bytes on any single HDD OSD.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 6,
|
|
"x": 6,
|
|
"y": 5
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "max((ceph_bluefs_slow_used_bytes{cluster=~\"$cluster\"} > 0) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"}) or vector(0)",
|
|
"refId": "A",
|
|
"legendFormat": ""
|
|
}
|
|
],
|
|
"options": {
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
],
|
|
"fields": "",
|
|
"values": false
|
|
},
|
|
"orientation": "auto",
|
|
"textMode": "auto",
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"justifyMode": "auto"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "bytes",
|
|
"decimals": null,
|
|
"mappings": [],
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
},
|
|
{
|
|
"color": "red",
|
|
"value": 1
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
}
|
|
},
|
|
{
|
|
"id": 5,
|
|
"type": "stat",
|
|
"title": "Max block.db utilization",
|
|
"description": "Fullest single HDD OSD's block.db. Note the volume selector exposes 122 GiB of the 128 GiB LV, so effective pressure is slightly higher than this raw ratio.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 6,
|
|
"x": 12,
|
|
"y": 5
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "max(100 * (ceph_bluefs_db_used_bytes{cluster=~\"$cluster\"} / ceph_bluefs_db_total_bytes{cluster=~\"$cluster\"}) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "A",
|
|
"legendFormat": ""
|
|
}
|
|
],
|
|
"options": {
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
],
|
|
"fields": "",
|
|
"values": false
|
|
},
|
|
"orientation": "auto",
|
|
"textMode": "auto",
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"justifyMode": "auto"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "percent",
|
|
"decimals": 1,
|
|
"mappings": [],
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
},
|
|
{
|
|
"color": "#EAB839",
|
|
"value": 70
|
|
},
|
|
{
|
|
"color": "red",
|
|
"value": 85
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
}
|
|
},
|
|
{
|
|
"id": 6,
|
|
"type": "stat",
|
|
"title": "Median block.db utilization",
|
|
"description": "Fleet median across HDD OSDs.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 6,
|
|
"x": 18,
|
|
"y": 5
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "quantile(0.5, 100 * (ceph_bluefs_db_used_bytes{cluster=~\"$cluster\"} / ceph_bluefs_db_total_bytes{cluster=~\"$cluster\"}) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "A",
|
|
"legendFormat": ""
|
|
}
|
|
],
|
|
"options": {
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
],
|
|
"fields": "",
|
|
"values": false
|
|
},
|
|
"orientation": "auto",
|
|
"textMode": "auto",
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"justifyMode": "auto"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "percent",
|
|
"decimals": 1,
|
|
"mappings": [],
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
},
|
|
{
|
|
"color": "#EAB839",
|
|
"value": 60
|
|
},
|
|
{
|
|
"color": "red",
|
|
"value": 80
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
}
|
|
},
|
|
{
|
|
"id": 7,
|
|
"type": "timeseries",
|
|
"title": "Spillover bytes by OSD",
|
|
"description": "Empty is healthy. A series appearing here is the signal to act.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 7,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 10
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "(ceph_bluefs_slow_used_bytes{cluster=~\"$cluster\"} > 0) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"}",
|
|
"refId": "A",
|
|
"legendFormat": "{{ceph_daemon}} @ {{hostname}}"
|
|
}
|
|
],
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom",
|
|
"showLegend": true,
|
|
"calcs": []
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "bytes",
|
|
"min": 0,
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"lineWidth": 2,
|
|
"fillOpacity": 18,
|
|
"showPoints": "never",
|
|
"spanNulls": true,
|
|
"axisLabel": "",
|
|
"gradientMode": "none"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"color": {
|
|
"mode": "palette-classic"
|
|
}
|
|
},
|
|
"overrides": []
|
|
}
|
|
},
|
|
{
|
|
"id": 8,
|
|
"type": "row",
|
|
"title": "block.db utilization",
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 17
|
|
},
|
|
"panels": []
|
|
},
|
|
{
|
|
"id": 9,
|
|
"type": "timeseries",
|
|
"title": "block.db utilization across HDD OSDs",
|
|
"description": "Spread matters more than the median: a widening max-to-median gap means metadata is landing unevenly, usually a hot bucket index or an unbalanced PG distribution.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 9,
|
|
"w": 16,
|
|
"x": 0,
|
|
"y": 18
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "max(100 * (ceph_bluefs_db_used_bytes{cluster=~\"$cluster\"} / ceph_bluefs_db_total_bytes{cluster=~\"$cluster\"}) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "A",
|
|
"legendFormat": "max"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "quantile(0.95, 100 * (ceph_bluefs_db_used_bytes{cluster=~\"$cluster\"} / ceph_bluefs_db_total_bytes{cluster=~\"$cluster\"}) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "B",
|
|
"legendFormat": "p95"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "quantile(0.5, 100 * (ceph_bluefs_db_used_bytes{cluster=~\"$cluster\"} / ceph_bluefs_db_total_bytes{cluster=~\"$cluster\"}) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "C",
|
|
"legendFormat": "median"
|
|
}
|
|
],
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom",
|
|
"showLegend": true,
|
|
"calcs": []
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "percent",
|
|
"min": 0,
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"lineWidth": 2,
|
|
"fillOpacity": 8,
|
|
"showPoints": "never",
|
|
"spanNulls": true,
|
|
"axisLabel": "",
|
|
"gradientMode": "none"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"color": {
|
|
"mode": "palette-classic"
|
|
},
|
|
"max": 100
|
|
},
|
|
"overrides": [
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "max"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#d95926"
|
|
}
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "p95"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#199e70"
|
|
}
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "median"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#3987e5"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
]
|
|
}
|
|
},
|
|
{
|
|
"id": 10,
|
|
"type": "bargauge",
|
|
"title": "Top 15 OSDs by block.db utilization",
|
|
"description": "Instant snapshot. Hostname is joined from ceph_osd_metadata.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 9,
|
|
"w": 8,
|
|
"x": 16,
|
|
"y": 18
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "topk(15, 100 * (ceph_bluefs_db_used_bytes{cluster=~\"$cluster\"} / ceph_bluefs_db_total_bytes{cluster=~\"$cluster\"}) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "A",
|
|
"legendFormat": "{{ceph_daemon}} {{hostname}}",
|
|
"instant": true
|
|
}
|
|
],
|
|
"options": {
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
],
|
|
"fields": "",
|
|
"values": false
|
|
},
|
|
"orientation": "horizontal",
|
|
"displayMode": "gradient",
|
|
"showUnfilled": true,
|
|
"minVizWidth": 8,
|
|
"minVizHeight": 10
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "percent",
|
|
"min": 0,
|
|
"mappings": [],
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
},
|
|
{
|
|
"color": "#EAB839",
|
|
"value": 70
|
|
},
|
|
{
|
|
"color": "red",
|
|
"value": 85
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
}
|
|
},
|
|
{
|
|
"id": 11,
|
|
"type": "row",
|
|
"title": "Growth and sizing",
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 27
|
|
},
|
|
"panels": []
|
|
},
|
|
{
|
|
"id": 12,
|
|
"type": "timeseries",
|
|
"title": "block.db used per OSD",
|
|
"description": "Absolute bytes rather than a ratio, so the growth rate is readable directly.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 12,
|
|
"x": 0,
|
|
"y": 28
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "max(ceph_bluefs_db_used_bytes{cluster=~\"$cluster\"} * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "A",
|
|
"legendFormat": "max"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "quantile(0.5, ceph_bluefs_db_used_bytes{cluster=~\"$cluster\"} * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "B",
|
|
"legendFormat": "median"
|
|
}
|
|
],
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom",
|
|
"showLegend": true,
|
|
"calcs": []
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "bytes",
|
|
"min": 0,
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"lineWidth": 2,
|
|
"fillOpacity": 8,
|
|
"showPoints": "never",
|
|
"spanNulls": true,
|
|
"axisLabel": "",
|
|
"gradientMode": "none"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"color": {
|
|
"mode": "palette-classic"
|
|
}
|
|
},
|
|
"overrides": [
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "max"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#d95926"
|
|
}
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "median"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#3987e5"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
]
|
|
}
|
|
},
|
|
{
|
|
"id": 13,
|
|
"type": "timeseries",
|
|
"title": "Cluster fill",
|
|
"description": "Deliberately a separate panel rather than a second axis on the block.db chart. CAVEAT: ceph_cluster_total_bytes counts each OSD's block.db LV inside raw HDD capacity, about 84 TiB fleet-wide, so this reads slightly low. Use pool metrics for capacity decisions; this panel is for correlating db growth against fill.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 12,
|
|
"x": 12,
|
|
"y": 28
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "100 * ceph_cluster_total_used_bytes{cluster=~\"$cluster\"} / ceph_cluster_total_bytes{cluster=~\"$cluster\"}",
|
|
"refId": "A",
|
|
"legendFormat": "raw used"
|
|
}
|
|
],
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom",
|
|
"showLegend": true,
|
|
"calcs": []
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "percent",
|
|
"min": 0,
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"lineWidth": 2,
|
|
"fillOpacity": 8,
|
|
"showPoints": "never",
|
|
"spanNulls": true,
|
|
"axisLabel": "",
|
|
"gradientMode": "none"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"color": {
|
|
"mode": "palette-classic"
|
|
},
|
|
"max": 100
|
|
},
|
|
"overrides": [
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "raw used"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#199e70"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
]
|
|
}
|
|
},
|
|
{
|
|
"id": 14,
|
|
"type": "timeseries",
|
|
"title": "block.db GiB per TiB stored on the OSD",
|
|
"description": "The sizing constant, derived live. If this climbs, average object size is falling and the crossover analysis needs redoing. It moves before spillover does.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 12,
|
|
"x": 0,
|
|
"y": 36
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "quantile(0.5, (ceph_bluefs_db_used_bytes{cluster=~\"$cluster\"} / 1073741824) / (ceph_osd_stat_bytes_used{cluster=~\"$cluster\"} / 1099511627776) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "A",
|
|
"legendFormat": "median"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "max((ceph_bluefs_db_used_bytes{cluster=~\"$cluster\"} / 1073741824) / (ceph_osd_stat_bytes_used{cluster=~\"$cluster\"} / 1099511627776) * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "B",
|
|
"legendFormat": "max"
|
|
}
|
|
],
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom",
|
|
"showLegend": true,
|
|
"calcs": []
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "none",
|
|
"min": 0,
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"lineWidth": 2,
|
|
"fillOpacity": 8,
|
|
"showPoints": "never",
|
|
"spanNulls": true,
|
|
"axisLabel": "",
|
|
"gradientMode": "none"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"color": {
|
|
"mode": "palette-classic"
|
|
}
|
|
},
|
|
"overrides": [
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "median"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#3987e5"
|
|
}
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "max"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#d95926"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
]
|
|
}
|
|
},
|
|
{
|
|
"id": 15,
|
|
"type": "timeseries",
|
|
"title": "BlueStore onode cache hit rate",
|
|
"description": "Fleet-wide onode cache effectiveness. Driven by osd_memory_target, which is unset on spice and therefore at the 4 GiB default. Raising it is the cheapest lever on small-object metadata performance and needs no OSD rebuild.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 12,
|
|
"x": 12,
|
|
"y": 36
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "100 * sum(rate(ceph_bluestore_onode_hits{cluster=~\"$cluster\"}[$__rate_interval])) / (sum(rate(ceph_bluestore_onode_hits{cluster=~\"$cluster\"}[$__rate_interval])) + sum(rate(ceph_bluestore_onode_misses{cluster=~\"$cluster\"}[$__rate_interval])))",
|
|
"refId": "A",
|
|
"legendFormat": "hit rate"
|
|
}
|
|
],
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom",
|
|
"showLegend": true,
|
|
"calcs": []
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "percent",
|
|
"min": 0,
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"lineWidth": 2,
|
|
"fillOpacity": 8,
|
|
"showPoints": "never",
|
|
"spanNulls": true,
|
|
"axisLabel": "",
|
|
"gradientMode": "none"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"color": {
|
|
"mode": "palette-classic"
|
|
},
|
|
"max": 100
|
|
},
|
|
"overrides": [
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "hit rate"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#199e70"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
]
|
|
}
|
|
},
|
|
{
|
|
"id": 16,
|
|
"type": "row",
|
|
"title": "BlueFS internals",
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 44
|
|
},
|
|
"panels": []
|
|
},
|
|
{
|
|
"id": 17,
|
|
"type": "timeseries",
|
|
"title": "BlueFS file count per OSD",
|
|
"description": "SST plus WAL file count. A step change usually means a compaction storm.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 7,
|
|
"w": 12,
|
|
"x": 0,
|
|
"y": 45
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "max(ceph_bluefs_num_files{cluster=~\"$cluster\"} * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "A",
|
|
"legendFormat": "max"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "quantile(0.5, ceph_bluefs_num_files{cluster=~\"$cluster\"} * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "B",
|
|
"legendFormat": "median"
|
|
}
|
|
],
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom",
|
|
"showLegend": true,
|
|
"calcs": []
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "none",
|
|
"min": 0,
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"lineWidth": 2,
|
|
"fillOpacity": 8,
|
|
"showPoints": "never",
|
|
"spanNulls": true,
|
|
"axisLabel": "",
|
|
"gradientMode": "none"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"color": {
|
|
"mode": "palette-classic"
|
|
}
|
|
},
|
|
"overrides": [
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "max"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#d95926"
|
|
}
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "median"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#3987e5"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
]
|
|
}
|
|
},
|
|
{
|
|
"id": 18,
|
|
"type": "timeseries",
|
|
"title": "BlueFS logged bytes per OSD",
|
|
"description": "BlueFS journal volume. Rides on the same NVMe as block.db and the OS.",
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"gridPos": {
|
|
"h": 7,
|
|
"w": 12,
|
|
"x": 12,
|
|
"y": 45
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "max(ceph_bluefs_logged_bytes{cluster=~\"$cluster\"} * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "A",
|
|
"legendFormat": "max"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "$datasource"
|
|
},
|
|
"expr": "quantile(0.5, ceph_bluefs_logged_bytes{cluster=~\"$cluster\"} * on(ceph_daemon) group_left(hostname) ceph_osd_metadata{cluster=~\"$cluster\", device_class=\"hdd\"})",
|
|
"refId": "B",
|
|
"legendFormat": "median"
|
|
}
|
|
],
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom",
|
|
"showLegend": true,
|
|
"calcs": []
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"unit": "bytes",
|
|
"min": 0,
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"lineWidth": 2,
|
|
"fillOpacity": 8,
|
|
"showPoints": "never",
|
|
"spanNulls": true,
|
|
"axisLabel": "",
|
|
"gradientMode": "none"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"color": {
|
|
"mode": "palette-classic"
|
|
}
|
|
},
|
|
"overrides": [
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "max"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#d95926"
|
|
}
|
|
}
|
|
]
|
|
},
|
|
{
|
|
"matcher": {
|
|
"id": "byName",
|
|
"options": "median"
|
|
},
|
|
"properties": [
|
|
{
|
|
"id": "color",
|
|
"value": {
|
|
"mode": "fixed",
|
|
"fixedColor": "#3987e5"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
]
|
|
}
|
|
}
|
|
]
|
|
} |