1、Pod.spec.nodeName 将 Pod 直接调度到指定的 Node 节点上，会跳过 Scheduler 的调度策略，该匹配规则是强制匹配
apiVersion: apps/v1 kind: Deployment metadata: name: myapp-pod6 spec: selector: matchLabels: app: myapp-pod6 replicas: 4 template: metadata: labels: app: myapp-pod6 spec: nodeName: xgcloud-ops-k8s-cluster-1.novalocal containers: - name: myapp-pod6 image: busybox command: [‘sh‘, ‘-c‘, ‘echo Hello Kubernetes! && sleep 3600‘]
kubectl apply -f pod6.yaml
kubectl get pod -o wide
2、Pod.spec.nodeSelector：通过 kubernetes 的 label-selector 机制选择节点，由调度器调度策略匹配 label，而后调度 Pod 到目标节点，该匹配规则属于强制约束
apiVersion: apps/v1 kind: Deployment metadata: name: myapp-pod7 spec: selector: matchLabels: app: myapp-pod7 replicas: 4 template: metadata: labels: app: myapp-pod7 spec: nodeSelector: kubernetes.io/hostname: xgcloud-ops-k8s-cluster-2.novalocal containers: - name: myapp-pod7 image: busybox command: [‘sh‘, ‘-c‘, ‘echo Hello Kubernetes! && sleep 3600‘]
kubectl apply -f pod7.yaml
kubectl get pod -o wide
原文：https://www.cnblogs.com/litzhiai/p/14978624.html