headless发现独立pod

如果需要让客户端连接到所有pod，需要找出每个pod的ip。可以通过dns查找发现pod ip，通常通过kube-dns查找服务名对应ip时只会显示服务的集群ip。但是如果此时不需要集群ip时，可以设置clusterIP为None，此时kube-dns返回的是每个pod ip而不是服务集群的ip。

headless：

apiVersion: v1
kind: Service
metadata:
  name: kubia-headless
spec:
  clusterIP: None
  ports:
  - port: 80
    targetPort: 80
  selector:
    app: kubia

deployment:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deployment
spec:
  selector:
    matchLabels:
      app: kubia
  replicas: 4 # tells deployment to run 2 pods matching the template
  template:
    metadata:
      labels:
        app: kubia
    spec:
      containers:
      - name: nginx
        image: nginx:1.14.2
        ports:
        - containerPort: 80

 

进入任意pod中进行域名解析：

root@nginx-test:/# nslookup kubia-headless
Server:         10.96.0.10
Address:        10.96.0.10#53

Name:   kubia-headless.default.svc.cluster.local
Address: 10.1.1.186
Name:   kubia-headless.default.svc.cluster.local
Address: 10.1.1.182
Name:   kubia-headless.default.svc.cluster.local
Address: 10.1.1.185
Name:   kubia-headless.default.svc.cluster.local
Address: 10.1.1.183

 

此时headless服务也是跨pod的负载均衡，但是是通过dns轮询机制而不是通过服务代理。