بدون KEDA:
  صف ۱۰۰٬۰۰۰ پیام → ۳ pod ثابت → backlog
  شب صف خالی → ۳ pod بیکار → هزینه
  HPA فقط روی CPU → صف پر ولی CPU پایین → اسکیل نمی‌شود

با KEDA:
  queue depth / lag / cron / Prometheus ──► ScaledObject
        │
        ├─ 0 → 1   (operator)
        └─ 1 → N   (HPA + external metrics)

 Event sources (Kafka, SQS, RabbitMQ, Prometheus, …)
              │ poll / push
              ▼
┌─────────────────────────────────────────────┐
│  keda-operator                              │
│   • watch ScaledObject / ScaledJob          │
│   • create/manage HPA                       │
│   • scale 0 ↔ 1 مستقیم                      │
├─────────────────────────────────────────────┤
│  keda-metrics-apiserver                     │
│   • external metrics → Kubernetes API       │
│   • HPA از اینجا برای 1 ↔ N می‌پرسد         │
├─────────────────────────────────────────────┤
│  keda-admission-webhooks                    │
│   • اعتبارسنجی ScaledObject هنگام apply     │
│   • مثلاً دو ScaledObject روی یک target     │
└─────────────────────────────────────────────┘
              │
              ▼
         HPA + ReplicaSet → Pods

apiVersion: keda.sh/v1alpha1
kind: ScaledObject
metadata:
  name: rabbitmq-consumer
  namespace: default
spec:
  scaleTargetRef:
    name: consumer-deploy          # Deployment / StatefulSet / ...
  pollingInterval: 30              # ثانیه — فاصلهٔ poll متریک
  cooldownPeriod: 300              # بعد از idle، قبل از scale-to-zero
  minReplicaCount: 0               # scale-to-zero
  maxReplicaCount: 50
  triggers:
    - type: rabbitmq
      metadata:
        protocol: amqp
        queueName: orders
        mode: QueueLength
        value: "5"                 # هدف: ~5 پیام per pod
      authenticationRef:
        name: rabbitmq-trigger-auth

apiVersion: keda.sh/v1alpha1
kind: ScaledObject
metadata:
  name: http-app-scaledobject
spec:
  scaleTargetRef:
    name: http-app
  minReplicaCount: 1
  maxReplicaCount: 10
  triggers:
    - type: prometheus
      metadata:
        serverAddress: http://prometheus-server.default.svc:9090
        threshold: "5"
        query: sum(rate(http_requests_total{app="http-app"}[1m]))

apiVersion: keda.sh/v1alpha1
kind: ScaledJob
metadata:
  name: image-processor
spec:
  jobTargetRef:
    template:
      spec:
        containers:
          - name: worker
            image: example/image-processor:1.0
        restartPolicy: Never
  pollingInterval: 30
  maxReplicaCount: 20
  successfulJobsHistoryLimit: 5
  failedJobsHistoryLimit: 5
  triggers:
    - type: aws-sqs-queue
      metadata:
        queueURL: https://sqs.eu-west-1.amazonaws.com/123/images
        queueLength: "1"
      authenticationRef:
        name: aws-trigger-auth

apiVersion: keda.sh/v1alpha1
kind: TriggerAuthentication
metadata:
  name: rabbitmq-trigger-auth
  namespace: default
spec:
  secretTargetRef:
    - parameter: host
      name: rabbitmq-secret
      key: host

helm repo add kedacore https://kedacore.github.io/charts
helm repo update
helm install keda kedacore/keda --namespace keda --create-namespace
kubectl get pods -n keda

# با admission webhooks — نسخه را با آخرین release عوض کنید
kubectl apply --server-side -f \
  https://github.com/kedacore/keda/releases/download/v2.20.0/keda-2.20.0.yaml

helm uninstall keda --namespace keda
# در صورت گیر کردن finalizer:
kubectl patch scaledobject <name> -p '{"metadata":{"finalizers":null}}' --type=merge

kubectl get scaledobject,scaledjob,hpa -A
kubectl logs -n keda deploy/keda-operator -f
kubectl get --raw "/apis/external.metrics.k8s.io/v1beta1" | head

Queue / Stream / Cron / Prometheus
        → KEDA (ScaledObject|ScaledJob)
        → HPA + Pods
        → (اختیاری) Cluster Autoscaler / Karpenter