Files
Johan LEROY dc952d13aa feat(monitoring): supervise l'API, la base et l'hôte avec Prometheus et Grafana
L'API exposait /metrics, mais aucun collecteur ne le lisait : monitoring/ ne contenait que des
.gitkeep.

Sous le profil Compose `monitoring` : prometheus, alertmanager, grafana, postgres-exporter,
node-exporter et cadvisor. Tous ont un mem_limit, pour environ 700 Mo au total sur la VM de 8 Go,
et leurs interfaces n'écoutent que sur 127.0.0.1. Le profil est actif en prod via
COMPOSE_PROFILES, donc à chaque déploiement, et se lance à la demande ailleurs
(make monitoring-up).

- Neuf règles d'alerte (API, base, hôte, cibles). Chacune a un cas dans les tests joués par
  `promtool test rules`, en CI comme par make monitoring-check.
- Alertmanager route les alertes par courriel vers Mailpit ; un critical masque le warning de la
  même cible.
- Grafana est provisionné : sources Prometheus et TimescaleDB, et trois tableaux de bord (API,
  données et dérive du modèle, infrastructure).
- Le rôle PostgreSQL `supervision` est en lecture seule sur les seules tables métier
  (db/roles/supervision.sql), posé par make db-ensure-supervision et par stack-up quand le
  profil est actif.
- Le jeton de /metrics passe à Prometheus en secret Compose (APP_METRICS_TOKEN) ;
  provision-host.sh génère ce secret et les deux autres.

Backend :
- un APP_METRICS_TOKEN vide vaut absent ;
- les sondes de santé ne comptent plus dans les métriques ;
- seaux de latence fins autour de 500 ms ;
- un registre Prometheus par application, sans quoi toute application créée après la première
  (dans les tests) ne mesurait rien.

Réf : #26
2026-09-23 09:41:12 +02:00

579 lines
13 KiB
JSON

{
"uid": "enervision-api",
"title": "EnerVision · API",
"description": "Débit, erreurs et latences vus par l'API elle-même (prometheus-fastapi-instrumentator).",
"tags": [
"enervision"
],
"timezone": "browser",
"editable": false,
"graphTooltip": 1,
"refresh": "10s",
"schemaVersion": 39,
"version": 1,
"time": {
"from": "now-1h",
"to": "now"
},
"templating": {
"list": []
},
"annotations": {
"list": []
},
"links": [],
"panels": [
{
"id": 1,
"type": "stat",
"title": "API",
"gridPos": {
"x": 0,
"y": 0,
"w": 6,
"h": 4
},
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"targets": [
{
"refId": "A",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "up{job=\"backend\"}",
"legendFormat": "",
"range": true
}
],
"fieldConfig": {
"defaults": {
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "red",
"value": null
},
{
"color": "green",
"value": 1
}
]
},
"color": {
"mode": "thresholds"
},
"mappings": [
{
"type": "value",
"options": {
"0": {
"text": "Injoignable",
"color": "red"
},
"1": {
"text": "Joignable",
"color": "green"
}
}
}
]
},
"overrides": []
},
"options": {
"reduceOptions": {
"calcs": [
"lastNotNull"
],
"fields": "",
"values": false
},
"colorMode": "background",
"graphMode": "area",
"textMode": "auto"
},
"description": "Prometheus joint-il /metrics ?"
},
{
"id": 2,
"type": "stat",
"title": "Débit",
"gridPos": {
"x": 6,
"y": 0,
"w": 6,
"h": 4
},
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"targets": [
{
"refId": "A",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "sum(rate(http_requests_total{job=\"backend\"}[5m]))",
"legendFormat": "",
"range": true
}
],
"fieldConfig": {
"defaults": {
"unit": "reqps",
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "blue",
"value": null
}
]
},
"color": {
"mode": "thresholds"
}
},
"overrides": []
},
"options": {
"reduceOptions": {
"calcs": [
"lastNotNull"
],
"fields": "",
"values": false
},
"colorMode": "background",
"graphMode": "area",
"textMode": "auto"
}
},
{
"id": 3,
"type": "stat",
"title": "Erreurs 5xx",
"gridPos": {
"x": 12,
"y": 0,
"w": 6,
"h": 4
},
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"targets": [
{
"refId": "A",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "sum(rate(http_requests_total{job=\"backend\", status=\"5xx\"}[5m])) / sum(rate(http_requests_total{job=\"backend\"}[5m]))",
"legendFormat": "",
"range": true
}
],
"fieldConfig": {
"defaults": {
"unit": "percentunit",
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": null
},
{
"color": "orange",
"value": 0.01
},
{
"color": "red",
"value": 0.05
}
]
},
"color": {
"mode": "thresholds"
}
},
"overrides": []
},
"options": {
"reduceOptions": {
"calcs": [
"lastNotNull"
],
"fields": "",
"values": false
},
"colorMode": "background",
"graphMode": "area",
"textMode": "auto"
},
"description": "Seuil d'alerte : 5 % pendant 5 minutes."
},
{
"id": 4,
"type": "stat",
"title": "Latence p95",
"gridPos": {
"x": 18,
"y": 0,
"w": 6,
"h": 4
},
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"targets": [
{
"refId": "A",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "histogram_quantile(0.95, sum by (le) (rate(http_request_duration_highr_seconds_bucket{job=\"backend\"}[5m])))",
"legendFormat": "",
"range": true
}
],
"fieldConfig": {
"defaults": {
"unit": "s",
"thresholds": {
"mode": "absolute",
"steps": [
{
"color": "green",
"value": null
},
{
"color": "orange",
"value": 0.5
},
{
"color": "red",
"value": 1
}
]
},
"color": {
"mode": "thresholds"
}
},
"overrides": []
},
"options": {
"reduceOptions": {
"calcs": [
"lastNotNull"
],
"fields": "",
"values": false
},
"colorMode": "background",
"graphMode": "area",
"textMode": "auto"
},
"description": "Seuil de charge : 500 ms (ADR 0015). Alerte au-delà d'une seconde pendant 10 minutes."
},
{
"id": 5,
"type": "timeseries",
"title": "Débit par route",
"gridPos": {
"x": 0,
"y": 4,
"w": 12,
"h": 8
},
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"targets": [
{
"refId": "A",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "sum by (handler) (rate(http_requests_total{job=\"backend\"}[1m]))",
"legendFormat": "{{handler}}",
"range": true
}
],
"fieldConfig": {
"defaults": {
"unit": "reqps"
},
"overrides": []
},
"options": {
"legend": {
"displayMode": "list",
"placement": "bottom",
"showLegend": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
}
},
{
"id": 6,
"type": "timeseries",
"title": "Latence de l'API",
"gridPos": {
"x": 12,
"y": 4,
"w": 12,
"h": 8
},
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"targets": [
{
"refId": "A",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "histogram_quantile(0.50, sum by (le) (rate(http_request_duration_highr_seconds_bucket{job=\"backend\"}[1m])))",
"legendFormat": "p50",
"range": true
},
{
"refId": "B",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "histogram_quantile(0.95, sum by (le) (rate(http_request_duration_highr_seconds_bucket{job=\"backend\"}[1m])))",
"legendFormat": "p95",
"range": true
},
{
"refId": "C",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "histogram_quantile(0.99, sum by (le) (rate(http_request_duration_highr_seconds_bucket{job=\"backend\"}[1m])))",
"legendFormat": "p99",
"range": true
}
],
"fieldConfig": {
"defaults": {
"unit": "s"
},
"overrides": []
},
"options": {
"legend": {
"displayMode": "list",
"placement": "bottom",
"showLegend": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
}
},
{
"id": 7,
"type": "timeseries",
"title": "Latence p95 par route",
"gridPos": {
"x": 0,
"y": 12,
"w": 12,
"h": 8
},
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"targets": [
{
"refId": "A",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "histogram_quantile(0.95, sum by (le, handler) (rate(http_request_duration_seconds_bucket{job=\"backend\"}[5m])))",
"legendFormat": "{{handler}}",
"range": true
}
],
"fieldConfig": {
"defaults": {
"unit": "s"
},
"overrides": []
},
"options": {
"legend": {
"displayMode": "list",
"placement": "bottom",
"showLegend": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
},
"description": "Estimée sur les seaux 50 ms à 2,5 s de http_request_duration_seconds."
},
{
"id": 8,
"type": "timeseries",
"title": "Réponses par classe de statut",
"gridPos": {
"x": 12,
"y": 12,
"w": 12,
"h": 8
},
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"targets": [
{
"refId": "A",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "sum by (status) (rate(http_requests_total{job=\"backend\"}[1m]))",
"legendFormat": "{{status}}",
"range": true
}
],
"fieldConfig": {
"defaults": {
"unit": "reqps"
},
"overrides": []
},
"options": {
"legend": {
"displayMode": "list",
"placement": "bottom",
"showLegend": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
}
},
{
"id": 9,
"type": "timeseries",
"title": "Mémoire du processus API",
"gridPos": {
"x": 0,
"y": 20,
"w": 12,
"h": 8
},
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"targets": [
{
"refId": "A",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "process_resident_memory_bytes{job=\"backend\"}",
"legendFormat": "résidente",
"range": true
}
],
"fieldConfig": {
"defaults": {
"unit": "bytes"
},
"overrides": []
},
"options": {
"legend": {
"displayMode": "list",
"placement": "bottom",
"showLegend": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
}
},
{
"id": 10,
"type": "timeseries",
"title": "Requêtes en cours de traitement",
"gridPos": {
"x": 12,
"y": 20,
"w": 12,
"h": 8
},
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"targets": [
{
"refId": "A",
"datasource": {
"type": "prometheus",
"uid": "prometheus"
},
"expr": "sum(rate(http_request_duration_highr_seconds_sum{job=\"backend\"}[1m]))",
"legendFormat": "secondes de traitement par seconde",
"range": true
}
],
"fieldConfig": {
"defaults": {
"unit": "short"
},
"overrides": []
},
"options": {
"legend": {
"displayMode": "list",
"placement": "bottom",
"showLegend": true
},
"tooltip": {
"mode": "multi",
"sort": "desc"
}
},
"description": "Loi de Little : durée moyenne multipliée par le débit, soit le nombre moyen de requêtes simultanées."
}
]
}