Revision history [back]

To configure a Kubernetes ingress to prioritize the local availability zone, you need to set the appropriate node selector labels for your nodes and your ingress resource. Here are the steps:

1. Label your Kubernetes nodes with the appropriate labels that indicate their availability zones. For example, if you have three availability zones (az-a, az-b, az-c), label your nodes accordingly:

kubectl label nodes node-a zone=az-a
kubectl label nodes node-b zone=az-b
kubectl label nodes node-c zone=az-c

1. When deploying your application pods, add a nodeSelector field to your PodSpec to ensure that they are deployed to a node in the local availability zone. For example:

apiVersion: v1
kind: Pod
metadata:
  name: my-app
spec:
  nodeSelector:
    zone: az-a # pod will be scheduled on a node in az-a
  containers:
  - name: my-container
    image: nginx

1. Label your ingress resources with the same nodeSelector field to ensure that traffic is routed to the local availability zone. For example:

apiVersion: networking.k8s.io/v1beta1
kind: Ingress
metadata:
  name: my-ingress
  annotations:
    nginx.ingress.kubernetes.io/affinity: "cookie"
spec:
  nodeSelector:
    zone: az-a # traffic will be routed to a node in az-a
  rules:
  - http:
      paths:
      - path: /testpath
        backend:
          serviceName: my-service
          servicePort: 80

With these configurations, Kubernetes will route traffic to the local availability zone first, ensuring better performance and reduced latency for your users.