# 模块 4：GPU Deployment — 生产级 GPU 工作负载
# 用法: kubectl apply -f 03_gpu_deploy.yaml
# 对应 PPT 第 45-46 页 [动手 1 + 2]
# 实验: kubectl scale deployment gpu-inference --replicas=3 → 观察资源争抢
apiVersion: apps/v1
kind: Deployment
metadata:
  name: gpu-inference
  namespace: gpu-demo
  labels:
    app: gpu-inference
spec:
  replicas: 1
  selector:
    matchLabels:
      app: gpu-inference
  template:
    metadata:
      labels:
        app: gpu-inference
    spec:
      containers:
      - name: cuda
        image: nvidia/cuda:12.4.0-base-ubuntu22.04
        command: ["/bin/bash", "-c"]
        args:
        - |
          echo "=== GPU Device Info ==="
          nvidia-smi
          echo "=== Environment ==="
          env | grep NVIDIA
          echo "GPU Pod is running. Sleeping..."
          sleep 3600
        resources:
          limits:
            nvidia.com/gpu: 1