apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: vllm-pdb
spec:
  minAvailable: 1
  selector:
    matchLabels:
      app: vllm-mock
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: vllm-mock
  labels:
    app: vllm-mock
spec:
  replicas: 2
  selector:
    matchLabels:
      app: vllm-mock
  template:
    metadata:
      labels:
        app: vllm-mock
    spec:
      terminationGracePeriodSeconds: 30
      # Prioritaskan Spot, fallback ke On-Demand jika Spot tidak tersedia
      affinity:
        nodeAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
          - weight: 100
            preference:
              matchExpressions:
              - key: node-type
                operator: In
                values:
                - spot
      containers:
      - name: mock-engine
        image: python:3.10-alpine
        command: ["python", "/app/server.py"]
        ports:
        - containerPort: 8000
        lifecycle:
          preStop:
            exec:
              command:
                # Beri jeda 5 detik agar kube-proxy/endpoints mencoret pod ini sebelum proses dimatikan
                - "/bin/sh"
                - "-c"
                - "sleep 5"
        readinessProbe:
          httpGet:
            path: /health
            port: 8000
          initialDelaySeconds: 2
          periodSeconds: 3
        volumeMounts:
        - name: script-vol
          mountPath: /app
      volumes:
      - name: script-vol
        configMap:
          name: mock-vllm-script