如何在GKE集群两节点池调度Pod？需确保默认池至少运行1个Pod

方案一：拆分Deployment（最稳妥） #

直接把4个Pod拆成两个独立的Deployment，彻底避免调度不确定性：

• 第一个Deployment固定1个副本，通过nodeAffinity强制调度到default-pool：

apiVersion: apps/v1
kind: Deployment
metadata:
  name: app-default
spec:
  replicas: 1
  selector:
    matchLabels:
      app: my-app
  template:
    metadata:
      labels:
        app: my-app
    spec:
      affinity:
        nodeAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            nodeSelectorTerms:
            - matchExpressions:
              - key: cloud.google.com/gke-nodepool
                operator: In
                values:
                - default-pool
      containers:
      - name: my-container
        image: your-image:tag

• 第二个Deployment设置3个副本，可灵活调度（也可配置优先调度到temp-pool）：

apiVersion: apps/v1
kind: Deployment
metadata:
  name: app-temp
spec:
  replicas: 3
  selector:
    matchLabels:
      app: my-app
  template:
    metadata:
      labels:
        app: my-app
    spec:
      affinity:
        nodeAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
          - weight: 100
            preference:
              matchExpressions:
              - key: cloud.google.com/gke-nodepool
                operator: In
                values:
                - temp-pool
      containers:
      - name: my-container
        image: your-image:tag

方案二：单Deployment结合拓扑分布约束+节点亲和性 #

如果不想拆分资源，可通过topologySpreadConstraints配合亲和性实现强制约束：

apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-app
spec:
  replicas: 4
  selector:
    matchLabels:
      app: my-app
  template:
    metadata:
      labels:
        app: my-app
    spec:
      affinity:
        nodeAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
          - weight: 50
            preference:
              matchExpressions:
              - key: cloud.google.com/gke-nodepool
                operator: In
                values:
                - default-pool
      topologySpreadConstraints:
      - maxSkew: 3
        topologyKey: cloud.google.com/gke-nodepool
        whenUnsatisfiable: DoNotSchedule
        labelSelector:
          matchLabels:
            app: my-app
      containers:
      - name: my-container
        image: your-image:tag

这里maxSkew:3限制两个节点池的Pod数量差不能超过3，总副本数为4时，调度器必须保证default-pool至少有1个Pod（否则temp-pool会有4个，差值为4，违反约束）。同时通过优先亲和性引导调度器优先往default-pool分配Pod。

注意事项 #

• GKE节点池默认标签为cloud.google.com/gke-nodepool，若自定义过节点标签，需替换为对应的键名
• 方案一逻辑简单、绝对可靠；方案二适合不想拆分资源的场景，但需确保拓扑分布参数配置准确

内容的提问来源于stack exchange，提问作者iamarunk