daemonset 和 service服务

1.nodeport 和ClusterIP

apiVersion: v1kind: Service#apiVersion: v1#kind: Servivemetadata:  name: nginx-svc  namespace: defaultspec:  type: NodePort  #定义类型为NodePort  selector:  #标签选择,把需要转发的后端的标签写在下面    app: nginx  ports:  - port: 80    #service端口    targetPort: 80    #容器端口    nodePort: 30080    #node端口，如果是ClusterIP的这段不需要，也可以不填会随机---apiVersion: v1kind: Podmetadata:  name: ng1  labels:    app: webspec:  containers:    - name: ng1      image:  hub.c.163.com/library/nginx      imagePullPolicy: IfNotPresent

nodeport 和ClusterIP 用法差不多就类型填的不一样，还有ClusterIP不用填nodePort

Deployment控制器

apiVersion: v1kind: Servicemetadata:name: nginx-svcnamespace: defaultspec:type: NodePort#定义类型为NodePortselector:#标签选择,把需要转发的后端的标签写在下面    app: webports:- port: 90    #service端口    targetPort: 80    #容器端口    nodePort: 30080    #node端口，如果是ClusterIP的这段不需要，也可以不填会随机---apiVersion: apps/v1kind: Deploymentmetadata:name: nginxspec:replicas: 2#表示启动两个容器selector:    matchLabels:            app: nginx#匹配标签template:#这下面的和pod写法一样    metadata:        labels:            app: nginx    spec:        containers:        - name: ng1            image:  hub.c.163.com/library/nginx            imagePullPolicy: IfNotPresent

2.daemonset控制器和service无头服务

apiVersion: apps/v1kind: DaemonSet#控制器类型dae...,这种控制器会把每一个node的节点都启动一个下面标签匹配的容器,有污点的除外metadata:  name: nginx-daemonset  namespace: defaultspec:  selector:    matchLabels:      app: nginx  template:    metadata:      labels:        app: nginx    spec:      containers:      - name: nginx        image: "ikubernetes/myapp:v1"---apiVersion: v1kind: Servicemetadata:  name: nginx-service  namespace: defaultspec:  selector:    app: nginx  clusterIP: None  ports:  - port: 80    targetPort: 80

service的无头服务适用于有些服务直接要和后端的容器建立连接场景使用的

[root@cs25 ~]# kubectl get pod  -l app=myapp -o wideNAME                            READY   STATUS    RESTARTS   AGE   IP               NODE   NOMINATED NODE   READINESS GATESmyapp-deploy-6c7db749dd-4kjdh   1/1     Running   0          31m   192.168.94.170   cs27              myapp-deploy-6c7db749dd-x95d6   1/1     Running   0          31m   192.168.100.2    cs25              #查看两个后端容器的ip[root@cs25 ~]# dig nginx-service.default.svc.cluster.local; <<>> DiG 9.9.4-RedHat-9.9.4-72.el7 <<>> nginx-service.default.svc.cluster.local;; global options: +cmd;; Got answer:;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 16814;; flags: qr aa rd; QUERY: 1, ANSWER: 2, AUTHORITY: 0, ADDITIONAL: 1;; WARNING: recursion requested but not available;; OPT PSEUDOSECTION:; EDNS: version: 0, flags:; udp: 4096;; QUESTION SECTION:;nginx-service.default.svc.cluster.local. IN A;; ANSWER SECTION:nginx-service.default.svc.cluster.local. 30 IN A 192.168.100.47nginx-service.default.svc.cluster.local. 30 IN A 192.168.94.168;; Query time: 4 msec;; SERVER: 192.168.100.37#53(192.168.100.37);; WHEN: 一 12月 23 14:03:19 CST 2019;; MSG SIZE  rcvd: 178#解析域名  nginx-service.default.svc.cluster.local 看是不是解析的容器ip，上面可以看到两个容器的ip都被解析到了，#如果这边是service带ip就会解析成service ip由service代理至后端容器，无头服务就是直接解析后端容器ip直接和容器建立连接。