NutriFlow — Configuracion Prometheus

# NutriFlow — Prometheus Configuration
# Cluster: GKE production · Region: eu-west1
# Generado por CULTIVA IA · configuracion-prometheus skill

global:
  scrape_interval:     15s
  evaluation_interval: 15s
  scrape_timeout:      10s
  external_labels:
    cluster:     "gke-nutriflow-prod"
    region:      "eu-west1"
    environment: "production"
    team:        "mlops"

alerting:
  alertmanagers:
    - static_configs:
        - targets:
            - "alertmanager.monitoring.svc:9093"
      timeout: 10s
      api_version: v2

rule_files:
  - "/etc/prometheus/rules/recording_rules.yml"
  - "/etc/prometheus/rules/alert_rules.yml"

scrape_configs:
  # --- Prometheus self-monitoring ---
  - job_name: "prometheus"
    static_configs:
      - targets: ["localhost:9090"]

  # --- Node exporters (3 nodos GKE) ---
  - job_name: "node-exporter"
    scrape_interval: 15s
    static_configs:
      - targets:
          - "gke-node-1.nutriflow.internal:9100"
          - "gke-node-2.nutriflow.internal:9100"
          - "gke-node-3.nutriflow.internal:9100"
        labels:
          env: "production"
          cluster: "gke-nutriflow-prod"
    relabel_configs:
      - source_labels: [__address__]
        target_label: instance
        regex: "([^:]+)(:[0-9]+)?"
        replacement: "${1}"

  # --- API Gateway (edge service) ---
  - job_name: "api-gateway"
    scrape_interval: 15s
    metrics_path: "/metrics"
    static_configs:
      - targets: ["api-gw.nutriflow.svc.cluster.local:8080"]
        labels:
          service: "api-gateway"
          tier:    "edge"
          slo:     "99.9"

  # --- Nutrition Engine (ML inference) — intervalo reducido ---
  - job_name: "nutrition-engine"
    scrape_interval: 10s  # Mas frecuente para inferencia critica
    metrics_path: "/metrics"
    static_configs:
      - targets: ["nutrition-engine.nutriflow.svc.cluster.local:8081"]
        labels:
          service: "nutrition-engine"
          tier:    "inference"
          model:   "nutrinet-v2"

  # --- Report Service ---
  - job_name: "report-service"
    scrape_interval: 15s
    static_configs:
      - targets: ["report-svc.nutriflow.svc.cluster.local:8082"]
        labels:
          service: "report-service"
          tier:    "backend"

  # --- PostgreSQL Exporter ---
  - job_name: "postgres-exporter"
    scrape_interval: 30s
    static_configs:
      - targets: ["postgres-exporter.nutriflow.svc.cluster.local:9187"]
        labels:
          db:      "postgresql"
          service: "nutriflow-db"

  # --- Redis Exporter ---
  - job_name: "redis-exporter"
    scrape_interval: 15s
    static_configs:
      - targets: ["redis-exporter.nutriflow.svc.cluster.local:9121"]
        labels:
          db:      "redis"
          service: "session-cache"

groups:
  # ── HTTP / API metrics ───────────────────────────────────────────
  - name: nutriflow_http_metrics
    interval: 15s
    rules:
      # Tasa de peticiones HTTP por servicio (req/s)
      - record: job:http_requests:rate5m
        expr: sum by (job) (rate(http_requests_total[5m]))

      # Tasa de errores 5xx
      - record: job:http_requests_errors:rate5m
        expr: sum by (job) (rate(http_requests_total{status=~"5.."}[5m]))

      # Porcentaje de error (usado en alertas)
      - record: job:http_error_rate:percent
        expr: |
            (job:http_requests_errors:rate5m / job:http_requests:rate5m) * 100

      # P95 latencia por servicio (principal SLO)
      - record: job:http_request_duration:p95
        expr: |
            histogram_quantile(0.95,
                sum by (job, le) (rate(http_request_duration_seconds_bucket[5m]))
              )

  # ── Resource / infrastructure metrics ───────────────────────────
  - name: nutriflow_resource_metrics
    interval: 30s
    rules:
      # CPU utilization % por nodo
      - record: instance:node_cpu:utilization
        expr: |
            100 - (avg by (instance) (
                rate(node_cpu_seconds_total{mode="idle"}[5m])
              ) * 100)

      # Memoria utilizada %
      - record: instance:node_memory:utilization
        expr: |
            100 - (
                (node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes) * 100
              )

      # Disco utilizado % (critico para report-service)
      - record: instance:node_disk:utilization
        expr: |
            100 - (
                (node_filesystem_avail_bytes{fstype!="tmpfs"}
                 / node_filesystem_size_bytes{fstype!="tmpfs"}) * 100
              )

groups:
  - name: nutriflow_availability
    interval: 30s
    rules:
      - alert: ServiceDown
        expr: up{job=~"api-gateway|nutrition-engine|report-service"} == 0
        for: 1m
        labels:
          severity: critical
          team:     mlops
          channel:  pagerduty
        annotations:
          summary:     "[CRITICO] Servicio {{ $labels.job }} caido"
          description: "{{ $labels.job }} en {{ $labels.instance }} lleva >1m sin responder"
          runbook_url: "https://runbook.nutriflow.io/service-down"

      - alert: MLInferenceLatency
        expr: job:http_request_duration:p95{job="nutrition-engine"} > 0.5
        for: 3m
        labels:
          severity: warning
          channel:  slack
        annotations:
          summary:     "Latencia ML degradada: {{ $value | humanizeDuration }}"
          description: "P95 inferencia nutrinet-v2 supera SLO de 500ms"