# КЛГ АСУ ТК — Prometheus Alerting Rules
groups:
  - name: klg-backend
    rules:
      - alert: HighErrorRate
        expr: rate(klg_http_errors_total[5m]) > 0.05
        for: 5m
        labels:
          severity: warning
        annotations:
          summary: "High error rate on KLG backend"
          description: "Error rate is {{ $value }}% over the last 5 minutes."

      - alert: SlowResponses
        expr: histogram_quantile(0.95, rate(klg_http_request_duration_seconds_bucket[5m])) > 2
        for: 10m
        labels:
          severity: warning
        annotations:
          summary: "Slow API responses"
          description: "95th percentile latency is {{ $value }}s."

      - alert: BackendDown
        expr: up{job="klg-backend"} == 0
        for: 2m
        labels:
          severity: critical
        annotations:
          summary: "KLG backend is down"
          description: "Backend has been unreachable for 2+ minutes."

      - alert: HighRequestRate
        expr: rate(klg_http_requests_total[1m]) > 100
        for: 5m
        labels:
          severity: info
        annotations:
          summary: "High request rate"
          description: "{{ $value }} requests/sec sustained for 5+ minutes."

      - alert: DatabaseErrors
        expr: increase(klg_http_errors_total{status="500"}[10m]) > 10
        for: 5m
        labels:
          severity: critical
        annotations:
          summary: "Multiple 500 errors"
          description: "{{ $value }} 500 errors in the last 10 minutes."