Will Anderson
bdba692db2
- Worker deployment: add COORDINATOR_URL and WORKER_ID env vars to portforward-helper; webhook calls coordinator /api/v2/workers/:id/port-lease after each NAT-PMP renewal so coordinator tracks current forwarded port - Worker overlays: patch WORKER_ID env var to actual worker ID (tx253/tx34) - Grafana: fornax-grafana Argo CD app pointing to k8s/fornax/grafana/ ConfigMap with grafana_dashboard=1 label for auto-discovery by sidecar; 7 panels covering bandwidth, active torrents, worker health, request rate, and error rate
305 lines
9.5 KiB
YAML
305 lines
9.5 KiB
YAML
apiVersion: v1
|
|
kind: ConfigMap
|
|
metadata:
|
|
name: fornax-grafana-dashboard
|
|
namespace: monitoring
|
|
labels:
|
|
grafana_dashboard: "1"
|
|
data:
|
|
fornax.json: |
|
|
{
|
|
"__inputs": [
|
|
{
|
|
"name": "DS_PROMETHEUS",
|
|
"label": "Prometheus",
|
|
"description": "",
|
|
"type": "datasource",
|
|
"pluginId": "prometheus",
|
|
"pluginName": "Prometheus"
|
|
},
|
|
{
|
|
"name": "DS_LOKI",
|
|
"label": "Loki",
|
|
"description": "",
|
|
"type": "datasource",
|
|
"pluginId": "loki",
|
|
"pluginName": "Loki"
|
|
}
|
|
],
|
|
"__requires": [
|
|
{ "type": "grafana", "id": "grafana", "name": "Grafana", "version": "9.0.0" },
|
|
{ "type": "datasource", "id": "prometheus", "name": "Prometheus", "version": "1.0.0" },
|
|
{ "type": "datasource", "id": "loki", "name": "Loki", "version": "1.0.0" }
|
|
],
|
|
"annotations": { "list": [] },
|
|
"description": "Fornax torrent distribution coordinator — fleet bandwidth, worker health, and distribution metrics",
|
|
"editable": true,
|
|
"fiscalYearStartMonth": 0,
|
|
"graphTooltip": 1,
|
|
"id": null,
|
|
"links": [],
|
|
"panels": [
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": { "h": 1, "w": 24, "x": 0, "y": 0 },
|
|
"id": 100,
|
|
"title": "Fleet Bandwidth",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": { "mode": "palette-classic" },
|
|
"custom": { "lineWidth": 2, "fillOpacity": 10 },
|
|
"unit": "Bps"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": { "h": 8, "w": 12, "x": 0, "y": 1 },
|
|
"id": 1,
|
|
"options": { "tooltip": { "mode": "multi" } },
|
|
"targets": [
|
|
{
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"expr": "sum(fornax_worker_download_bytes_per_second)",
|
|
"legendFormat": "Total Download",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Total Download Bandwidth",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": { "mode": "palette-classic" },
|
|
"custom": { "lineWidth": 2, "fillOpacity": 10 },
|
|
"unit": "Bps"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": { "h": 8, "w": 12, "x": 12, "y": 1 },
|
|
"id": 2,
|
|
"options": { "tooltip": { "mode": "multi" } },
|
|
"targets": [
|
|
{
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"expr": "sum(fornax_worker_upload_bytes_per_second)",
|
|
"legendFormat": "Total Upload",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Total Upload Bandwidth",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": { "mode": "palette-classic" },
|
|
"custom": { "lineWidth": 2, "fillOpacity": 10 },
|
|
"unit": "Bps"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": { "h": 8, "w": 24, "x": 0, "y": 9 },
|
|
"id": 3,
|
|
"options": { "tooltip": { "mode": "multi" } },
|
|
"targets": [
|
|
{
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"expr": "fornax_worker_download_bytes_per_second",
|
|
"legendFormat": "Worker {{ worker_id }}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Per-Worker Download Bandwidth",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": { "h": 1, "w": 24, "x": 0, "y": 17 },
|
|
"id": 101,
|
|
"title": "Torrent Distribution",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": { "mode": "palette-classic" },
|
|
"unit": "short",
|
|
"min": 0
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": { "h": 8, "w": 12, "x": 0, "y": 18 },
|
|
"id": 4,
|
|
"options": {
|
|
"orientation": "horizontal",
|
|
"reduceOptions": { "calcs": ["lastNotNull"] },
|
|
"text": {}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"expr": "fornax_worker_active_torrents",
|
|
"legendFormat": "{{ worker_id }}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Active Torrent Count per Worker",
|
|
"type": "bargauge"
|
|
},
|
|
{
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": { "mode": "thresholds" },
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{ "color": "red", "value": null },
|
|
{ "color": "yellow", "value": 1 },
|
|
{ "color": "green", "value": 1 }
|
|
]
|
|
},
|
|
"mappings": [
|
|
{ "options": { "0": { "color": "red", "text": "DOWN" }, "1": { "color": "green", "text": "UP" } }, "type": "value" }
|
|
]
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": { "h": 8, "w": 12, "x": 12, "y": 18 },
|
|
"id": 5,
|
|
"options": {
|
|
"footer": { "show": false },
|
|
"showHeader": true
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": { "type": "prometheus", "uid": "${datasource}" },
|
|
"expr": "fornax_worker_healthy",
|
|
"format": "table",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Worker Health Status",
|
|
"transformations": [
|
|
{
|
|
"id": "organize",
|
|
"options": {
|
|
"renameByName": {
|
|
"worker_id": "Worker ID",
|
|
"Value": "Status",
|
|
"Time": "Last Seen"
|
|
}
|
|
}
|
|
}
|
|
],
|
|
"type": "table"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": { "h": 1, "w": 24, "x": 0, "y": 26 },
|
|
"id": 102,
|
|
"title": "Coordinator Logs & Errors",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": { "type": "loki", "uid": "${loki}" },
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": { "mode": "palette-classic" },
|
|
"custom": { "lineWidth": 2, "fillOpacity": 10 },
|
|
"unit": "reqps"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": { "h": 8, "w": 12, "x": 0, "y": 27 },
|
|
"id": 6,
|
|
"options": { "tooltip": { "mode": "multi" } },
|
|
"targets": [
|
|
{
|
|
"datasource": { "type": "loki", "uid": "${loki}" },
|
|
"expr": "rate({app=\"fornax-coordinator\"} |= \"POST /api/v2\" [1m])",
|
|
"legendFormat": "Request Rate",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Coordinator Request Rate",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": { "type": "loki", "uid": "${loki}" },
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": { "mode": "palette-classic" },
|
|
"custom": { "lineWidth": 2, "fillOpacity": 10 },
|
|
"unit": "reqps",
|
|
"color": { "fixedColor": "red", "mode": "fixed" }
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": { "h": 8, "w": 12, "x": 12, "y": 27 },
|
|
"id": 7,
|
|
"options": { "tooltip": { "mode": "multi" } },
|
|
"targets": [
|
|
{
|
|
"datasource": { "type": "loki", "uid": "${loki}" },
|
|
"expr": "rate({app=\"fornax-coordinator\"} |= \"level=error\" [5m])",
|
|
"legendFormat": "Error Rate",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Error Rate",
|
|
"type": "timeseries"
|
|
}
|
|
],
|
|
"refresh": "30s",
|
|
"schemaVersion": 37,
|
|
"style": "dark",
|
|
"tags": ["fornax", "torrents", "monitoring"],
|
|
"templating": {
|
|
"list": [
|
|
{
|
|
"current": {},
|
|
"hide": 0,
|
|
"includeAll": false,
|
|
"label": "Prometheus",
|
|
"multi": false,
|
|
"name": "datasource",
|
|
"options": [],
|
|
"query": "prometheus",
|
|
"refresh": 1,
|
|
"regex": "",
|
|
"type": "datasource"
|
|
},
|
|
{
|
|
"current": {},
|
|
"hide": 0,
|
|
"includeAll": false,
|
|
"label": "Loki",
|
|
"multi": false,
|
|
"name": "loki",
|
|
"options": [],
|
|
"query": "loki",
|
|
"refresh": 1,
|
|
"regex": "",
|
|
"type": "datasource"
|
|
}
|
|
]
|
|
},
|
|
"time": { "from": "now-3h", "to": "now" },
|
|
"timepicker": {},
|
|
"timezone": "browser",
|
|
"title": "Fornax — Fleet Overview",
|
|
"uid": "fornax-fleet-overview",
|
|
"version": 1,
|
|
"weekStart": ""
|
|
}
|