apiVersion: monitoring.coreos.com/v1
kind: PrometheusRule
metadata: {name: cloud-command-slo, namespace: observability}
spec:
  groups:
    - name: cloud-command
      rules:
        - alert: CloudCommandHighErrorRate
          expr: sum(rate(http_requests_total{service="cloud-command",status=~"5.."}[5m])) / sum(rate(http_requests_total{service="cloud-command"}[5m])) > 0.02
          for: 10m
          labels: {severity: critical, service: cloud-command}
          annotations: {summary: "Cloud Command error rate exceeds 2%"}
        - alert: CloudCommandHpaSaturated
          expr: kube_horizontalpodautoscaler_status_current_replicas{horizontalpodautoscaler="cloud-command-api"} >= kube_horizontalpodautoscaler_spec_max_replicas
          for: 15m
          labels: {severity: warning, service: cloud-command}