apiVersion: v1
kind: ServiceAccount
metadata:
  name: spot-handler-sa
  namespace: default
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: spot-handler-role
rules:
  # Izin untuk cordon node
  - apiGroups: [""]
    resources: ["nodes"]
    verbs: ["get", "patch", "update"]
  # Izin untuk membaca dan mengevakuasi pod
  - apiGroups: [""]
    resources: ["pods"]
    verbs: ["get", "list", "watch"]
  - apiGroups: [""]
    resources: ["pods/eviction"]
    verbs: ["create"]
  - apiGroups: ["apps"]
    resources: ["daemonsets"]
    verbs: ["get"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: spot-handler-binding
subjects:
  - kind: ServiceAccount
    name: spot-handler-sa
    namespace: default
roleRef:
  kind: ClusterRole
  name: spot-handler-role
  apiGroup: rbac.authorization.k8s.io
---
apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: spot-termination-handler
  namespace: default
spec:
  selector:
    matchLabels:
      app: spot-termination-handler
  template:
    metadata:
      labels:
        app: spot-termination-handler
    spec:
      serviceAccountName: spot-handler-sa
      # Hanya deploy ke node yang memiliki label node-type: spot
      nodeSelector:
        node-type: spot
      containers:
      - name: handler
        image: bitnami/kubectl:latest
        env:
        # Mengambil nama node tempat pod ini berjalan secara otomatis
        - name: NODE_NAME
          valueFrom:
            fieldRef:
              fieldPath: spec.nodeName
        - name: TRIGGER_FILE
          value: "/tmp/preempt_signal"
        command:
        - "/bin/sh"
        - "-c"
        - |
          echo "[INFO] Spot Termination Handler active on node: $NODE_NAME"
          echo "[INFO] Listening for termination event..."
          
          # Di cloud nyata: curl ke http://169.254.169.254/... (AWS/GCP metadata)
          # Di lab Kind: kita monitor file trigger /tmp/preempt_signal
          while true; do
            if [ -f "$TRIGGER_FILE" ]; then
              echo "[ALERT] Preemption signal received on $NODE_NAME!"
              echo "[ACTION] Cordoning node $NODE_NAME..."
              kubectl cordon "$NODE_NAME"
              
              echo "[ACTION] Draining node $NODE_NAME..."
              kubectl drain "$NODE_NAME" \
                --ignore-daemonsets \
                --delete-emptydir-data \
                --grace-period=30 \
                --force
                
              echo "[DONE] Node $NODE_NAME safely evacuated. Awaiting cloud teardown."
              rm -f "$TRIGGER_FILE"
              break
            fi
            sleep 2
          done
          
          # Jaga container tetap hidup agar log bisa diperiksa
          sleep 3600