Monitoring avec Prometheus et Grafana

Un serveur sans monitoring, c'est voler à l'aveugle. Prometheus + Grafana est le standard de facto open source pour surveiller votre infrastructure.

Architecture

Applications / Serveurs
        ↓ métriques (pull)
   Prometheus (collecte + stockage)
        ↓ requêtes PromQL
    Grafana (visualisation)
        ↓ alertes
  AlertManager → Slack / Email

Installation avec Docker Compose

# docker-compose.monitoring.yml
version: '3.9'

services:
  prometheus:
    image: prom/prometheus:latest
    volumes:
      - ./prometheus.yml:/etc/prometheus/prometheus.yml
      - prometheus-data:/prometheus
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.retention.time=30d'
    ports:
      - "9090:9090"

  grafana:
    image: grafana/grafana:latest
    environment:
      - GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_PASSWORD}
      - GF_USERS_ALLOW_SIGN_UP=false
    volumes:
      - grafana-data:/var/lib/grafana
    ports:
      - "3000:3000"
    depends_on:
      - prometheus

  node-exporter:
    image: prom/node-exporter:latest
    volumes:
      - /proc:/host/proc:ro
      - /sys:/host/sys:ro
      - /:/rootfs:ro
    command:
      - '--path.procfs=/host/proc'
      - '--path.sysfs=/host/sys'
      - '--collector.filesystem.ignored-mount-points=^/(sys|proc|dev|host|etc)($$|/)'
    ports:
      - "9100:9100"

  cadvisor:
    image: gcr.io/cadvisor/cadvisor:latest
    volumes:
      - /:/rootfs:ro
      - /var/run:/var/run:rw
      - /sys:/sys:ro
      - /var/lib/docker/:/var/lib/docker:ro
    ports:
      - "8080:8080"

volumes:
  prometheus-data:
  grafana-data:

Configuration Prometheus

# prometheus.yml
global:
  scrape_interval: 15s
  evaluation_interval: 15s

alerting:
  alertmanagers:
    - static_configs:
        - targets: ['alertmanager:9093']

rule_files:
  - "alerts/*.yml"

scrape_configs:
  - job_name: 'prometheus'
    static_configs:
      - targets: ['localhost:9090']

  - job_name: 'node-exporter'
    static_configs:
      - targets: ['node-exporter:9100']
    relabel_configs:
      - source_labels: [__address__]
        target_label: instance

  - job_name: 'cadvisor'
    static_configs:
      - targets: ['cadvisor:8080']

  - job_name: 'laravel'
    static_configs:
      - targets: ['app:9001']
    metrics_path: '/metrics'

Métriques Laravel avec Prometheus

composer require spatie/laravel-prometheus
php artisan vendor:publish --provider="Spatie\LaravelPrometheus\LaravelPrometheusServiceProvider"

// Métriques personnalisées
use Spatie\LaravelPrometheus\Metrics\Counter;
use Spatie\LaravelPrometheus\Metrics\Histogram;

// Dans un middleware ou service
Counter::name('http_requests_total')
    ->help('Total HTTP requests')
    ->labels(['method', 'route', 'status'])
    ->register()
    ->increment(['GET', 'articles.index', '200']);

Histogram::name('http_request_duration_seconds')
    ->help('HTTP request duration')
    ->buckets([0.1, 0.25, 0.5, 1, 2.5, 5, 10])
    ->register()
    ->observe(0.234, ['GET', 'articles.index']);

Requêtes PromQL essentielles

# CPU usage (%)
100 - (avg by (instance) (rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)

# RAM disponible
node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes * 100

# Disk usage (%)
(node_filesystem_size_bytes - node_filesystem_avail_bytes)
    / node_filesystem_size_bytes * 100

# Requêtes HTTP par seconde
rate(http_requests_total[5m])

# Latence P95
histogram_quantile(0.95, rate(http_request_duration_seconds_bucket[5m]))

# Containers en cours d'exécution
count(container_last_seen{name!=""} > time() - 60)

Alertes

# alerts/server.yml
groups:
  - name: server
    rules:
      - alert: HighCPU
        expr: 100 - (avg by (instance) (rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100) > 85
        for: 5m
        labels:
          severity: warning
        annotations:
          summary: "CPU élevé sur {{ $labels.instance }}"
          description: "CPU à {{ $value | humanize }}%"

      - alert: DiskAlmostFull
        expr: (node_filesystem_size_bytes - node_filesystem_avail_bytes) / node_filesystem_size_bytes * 100 > 85
        for: 1m
        labels:
          severity: critical
        annotations:
          summary: "Disque presque plein sur {{ $labels.instance }}"

Dashboards Grafana recommandés

Importez ces dashboards depuis grafana.com :

• 1860 — Node Exporter Full (métriques système complètes)
• 893 — Docker and system monitoring
• 14057 — Laravel metrics

Accédez à Grafana sur http://votre-serveur:3000 (admin / votre password) et importez via Dashboards → Import → ID.