groups:
  - name: backup
    rules:
      - alert: BackupJobFailed
        expr: increase(manage_backup_runs_total{status="failed"}[1h]) > 0
        for: 0m
        labels:
          severity: critical
          category: backup
        annotations:
          summary: "Backup job {{ $labels.job_name }} failed"
          description: "At least one failed backup run was recorded for job {{ $labels.job_name }} within the last hour."

      - alert: BackupJobStuck
        expr: time() - manage_backup_runs_last_success_timestamp > 86400
        for: 5m
        labels:
          severity: warning
          category: backup
        annotations:
          summary: "Backup job {{ $labels.job_name }} has not succeeded in 24h"
          description: "No successful backup run has been recorded for job {{ $labels.job_name }} in the last 24 hours."

  - name: observability_health
    rules:
      - alert: PrometheusTargetMissing
        expr: up == 0
        for: 2m
        labels:
          severity: warning
          category: observability
        annotations:
          summary: "Prometheus target {{ $labels.job }} / {{ $labels.instance }} is down"
          description: "The Prometheus scrape target for {{ $labels.job }} on {{ $labels.instance }} has been unreachable for more than 2 minutes."

      - alert: AlertmanagerDown
        expr: up{job="alertmanager"} == 0
        for: 2m
        labels:
          severity: critical
          category: observability
        annotations:
          summary: "Alertmanager is down"
          description: "Alertmanager has been unreachable for more than 2 minutes. Alerts may not be delivered."

      - alert: GrafanaDown
        expr: up{job="grafana"} == 0
        for: 2m
        labels:
          severity: warning
          category: observability
        annotations:
          summary: "Grafana is down"
          description: "Grafana has been unreachable for more than 2 minutes. Dashboards are unavailable."