Kubernetes K8S之Pod跨namespace名称空间访问Service服务

Kubernetes的两个Service（ServiceA、ServiceB）和对应的Pod（PodA、PodB）分别属于不同的namespace名称空间，现需要PodA和PodB跨namespace名称空间并通过Service实现互访。应该如何实现？

场景需求

Kubernetes的两个Service（ServiceA、ServiceB）和对应的Pod（PodA、PodB）分别属于不同的namespace名称空间，现需要PodA和PodB跨namespace名称空间并通过Service实现互访。如何实现？

说明：这里是指通过Service的Name进行通信访问，而不是通过Service的IP【因因为每次重启Service，NAME不会改变，而IP是会改变的】。

主机配置规划

创建Service和Pod

相关yaml文件

[root@k8s-master cross_ns]# pwd  /root/k8s_practice/cross_ns  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# cat deply_service_myns.yaml  apiVersion: v1  kind: Namespace  metadata:    name: myns  ---  apiVersion: apps/v1  kind: Deployment  metadata:    name: myapp-deploy1    namespace: myns  spec:    replicas: 2    selector:      matchLabels:        app: myapp        release: v1    template:      metadata:        labels:          app: myapp          release: v1      spec:        containers:        - name: myapp          image: registry.cn-beijing.aliyuncs.com/google_registry/myapp:v1          imagePullPolicy: IfNotPresent          ports:          - name: http            containerPort: 80  ---  apiVersion: v1  kind: Service  metadata:    name: myapp-clusterip1    namespace: myns  spec:    type: ClusterIP  # 默认类型    selector:      app: myapp      release: v1    ports:    - name: http      port: 80      targetPort: 80    [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# cat deply_service_mytest.yaml  apiVersion: v1  kind: Namespace  metadata:    name: mytest  ---  apiVersion: apps/v1  kind: Deployment  metadata:    name: myapp-deploy2    namespace: mytest  spec:    replicas: 2    selector:      matchLabels:        app: myapp        release: v2    template:      metadata:        labels:          app: myapp          release: v2      spec:        containers:        - name: myapp          image: registry.cn-beijing.aliyuncs.com/google_registry/myapp:v2          imagePullPolicy: IfNotPresent          ports:          - name: http            containerPort: 80  ---  apiVersion: v1  kind: Service  metadata:    name: myapp-clusterip2    namespace: mytest  spec:    type: ClusterIP  # 默认类型    selector:      app: myapp      release: v2    ports:    - name: http      port: 80      targetPort: 80

运行yaml文件

kubectl apply -f deply_service_myns.yaml  kubectl apply -f deply_service_mytest.yaml

查看myns名称空间信息

[root@k8s-master cross_ns]# kubectl get svc -n myns -o wide  NAME               TYPE        CLUSTER-IP     EXTERNAL-IP   PORT(S)   AGE   SELECTOR  myapp-clusterip1   ClusterIP   10.100.61.11   <none>        80/TCP    3m    app=myapp,release=v1  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# kubectl get deploy -n myns -o wide  NAME            READY   UP-TO-DATE   AVAILABLE   AGE    CONTAINERS   IMAGES                                                      SELECTOR  myapp-deploy1   2/2     2            2           3m7s   myapp        registry.cn-beijing.aliyuncs.com/google_registry/myapp:v1   app=myapp,release=v1  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# kubectl get rs -n myns -o wide  NAME                       DESIRED   CURRENT   READY   AGE     CONTAINERS   IMAGES                                                      SELECTOR  myapp-deploy1-5b9d78576c   2         2         2       3m15s   myapp        registry.cn-beijing.aliyuncs.com/google_registry/myapp:v1   app=myapp,pod-template-hash=5b9d78576c,release=v1  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# kubectl get pod -n myns -o wide  NAME                             READY   STATUS    RESTARTS   AGE     IP             NODE         NOMINATED NODE   READINESS GATES  myapp-deploy1-5b9d78576c-wfw4n   1/1     Running   0          3m20s   10.244.2.136   k8s-node02   <none>           <none>  myapp-deploy1-5b9d78576c-zsfjl   1/1     Running   0          3m20s   10.244.3.193   k8s-node01   <none>           <none>

查看mytest名称空间信息

[root@k8s-master cross_ns]# kubectl get svc -n mytest -o wide  NAME               TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)   AGE    SELECTOR  myapp-clusterip2   ClusterIP   10.100.201.103   <none>        80/TCP    4m9s   app=myapp,release=v2  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# kubectl get deploy -n mytest -o wide  NAME            READY   UP-TO-DATE   AVAILABLE   AGE     CONTAINERS   IMAGES                                                      SELECTOR  myapp-deploy2   2/2     2            2           4m15s   myapp        registry.cn-beijing.aliyuncs.com/google_registry/myapp:v2   app=myapp,release=v2  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# kubectl get rs -n mytest -o wide  NAME                      DESIRED   CURRENT   READY   AGE     CONTAINERS   IMAGES                                                      SELECTOR  myapp-deploy2-dc8f96497   2         2         2       4m22s   myapp        registry.cn-beijing.aliyuncs.com/google_registry/myapp:v2   app=myapp,pod-template-hash=dc8f96497,release=v2  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# kubectl get pod -n mytest -o wide  NAME                            READY   STATUS    RESTARTS   AGE     IP             NODE         NOMINATED NODE   READINESS GATES  myapp-deploy2-dc8f96497-nnkqn   1/1     Running   0          4m27s   10.244.3.194   k8s-node01   <none>           <none>  myapp-deploy2-dc8f96497-w47dt   1/1     Running   0          4m27s   10.244.2.137   k8s-node02   <none>           <none>

只看Service和Pod

[root@k8s-master cross_ns]# kubectl get pod -A -o wide | grep -E '(my)|(NAME)'  NAMESPACE              NAME                                         READY   STATUS    RESTARTS   AGE   IP             NODE         NOMINATED NODE   READINESS GATES  myns                   myapp-deploy1-5b9d78576c-wfw4n               1/1     Running   0          41m   10.244.2.136   k8s-node02   <none>           <none>  myns                   myapp-deploy1-5b9d78576c-zsfjl               1/1     Running   0          41m   10.244.3.193   k8s-node01   <none>           <none>  mytest                 myapp-deploy2-dc8f96497-nnkqn                1/1     Running   0          41m   10.244.3.194   k8s-node01   <none>           <none>  mytest                 myapp-deploy2-dc8f96497-w47dt                1/1     Running   0          41m   10.244.2.137   k8s-node02   <none>           <none>  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# kubectl get svc -A -o wide | grep -E '(my)|(NAME)'  NAMESPACE              NAME                        TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                  AGE   SELECTOR  myns                   myapp-clusterip1            ClusterIP   10.100.61.11     <none>        80/TCP                   41m   app=myapp,release=v1  mytest                 myapp-clusterip2            ClusterIP   10.100.201.103   <none>        80/TCP                   41m   app=myapp,release=v2

pod跨名称空间namespace与Service通信

说明：是通过Service的NAME进行通信，而不是Service的IP【因为每次重启Service，NAME不会改变，而IP是会改变的】。

# 进入ns名称空间下的一个Pod容器  [root@k8s-master cross_ns]# kubectl exec -it -n myns myapp-deploy1-5b9d78576c-wfw4n sh  / # cd /root/  ### 如下说明在同一名称空间下，通信无问题  ~ # ping myapp-clusterip1  PING myapp-clusterip1 (10.100.61.11): 56 data bytes  64 bytes from 10.100.61.11: seq=0 ttl=64 time=0.046 ms  64 bytes from 10.100.61.11: seq=1 ttl=64 time=0.081 ms  ~ #  ~ # wget myapp-clusterip1 -O myns.html  Connecting to myapp-clusterip1 (10.100.61.11:80)  myns.html            100%  ~ #  ~ # cat myns.html  Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>    ### 如下说明在不同的名称空间下，通过Service的NAME进行通信存在问题  ~ # ping myapp-clusterip2  ping: bad address 'myapp-clusterip2'  ~ #  ~ # wget myapp-clusterip2 -O mytest.html  wget: bad address 'myapp-clusterip2'

实现跨namespace与Service通信

通过Service的ExternalName类型即可实现跨namespace名称空间与Service通信。

Service域名格式：$(service name).$(namespace).svc.cluster.local，其中 cluster.local 为指定的集群的域名

相关yaml文件

[root@k8s-master cross_ns]# pwd  /root/k8s_practice/cross_ns  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# cat svc_ExternalName_visit.yaml  # 实现 myns 名称空间的pod，访问 mytest 名称空间的Service：myapp-clusterip2  apiVersion: v1  kind: Service  metadata:    name: myapp-clusterip1-externalname    namespace: myns  spec:    type: ExternalName    externalName: myapp-clusterip2.mytest.svc.cluster.local    ports:    - name: http      port: 80      targetPort: 80  ---  # 实现 mytest 名称空间的Pod，访问 myns 名称空间的Service：myapp-clusterip1  apiVersion: v1  kind: Service  metadata:    name: myapp-clusterip2-externalname    namespace: mytest  spec:    type: ExternalName    externalName: myapp-clusterip1.myns.svc.cluster.local    ports:    - name: http      port: 80      targetPort: 80

运行yaml文件

[root@k8s-master cross_ns]# kubectl apply -f svc_ExternalName_visit.yaml  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# kubectl get svc -A -o wide | grep -E '(ExternalName)|(NAME)'  NAMESPACE              NAME                            TYPE           CLUSTER-IP       EXTERNAL-IP                                 PORT(S)                  AGE   SELECTOR  myns                   myapp-clusterip1-externalname   ExternalName   <none>           myapp-clusterip2.mytest.svc.cluster.local   80/TCP                   28s   <none>  mytest                 myapp-clusterip2-externalname   ExternalName   <none>           myapp-clusterip1.myns.svc.cluster.local     80/TCP                   28s   <none>

pod跨名称空间namespace与Service通信

到目前所有service和pod信息查看

[root@k8s-master cross_ns]# kubectl get svc -A -o wide | grep -E '(my)|(NAME)'  NAMESPACE              NAME                            TYPE           CLUSTER-IP       EXTERNAL-IP                                 PORT(S)                  AGE   SELECTOR  myns                   myapp-clusterip1                ClusterIP      10.100.61.11     <none>                                      80/TCP                   62m   app=myapp,release=v1  myns                   myapp-clusterip1-externalname   ExternalName   <none>           myapp-clusterip2.mytest.svc.cluster.local   80/TCP                   84s   <none>  mytest                 myapp-clusterip2                ClusterIP      10.100.201.103   <none>                                      80/TCP                   62m   app=myapp,release=v2  mytest                 myapp-clusterip2-externalname   ExternalName   <none>           myapp-clusterip1.myns.svc.cluster.local     80/TCP                   84s   <none>  [root@k8s-master cross_ns]#  [root@k8s-master cross_ns]# kubectl get pod -A -o wide | grep -E '(my)|(NAME)'  NAMESPACE              NAME                                         READY   STATUS    RESTARTS   AGE   IP             NODE         NOMINATED NODE   READINESS GATES  myns                   myapp-deploy1-5b9d78576c-wfw4n               1/1     Running   0          62m   10.244.2.136   k8s-node02   <none>           <none>  myns                   myapp-deploy1-5b9d78576c-zsfjl               1/1     Running   0          62m   10.244.3.193   k8s-node01   <none>           <none>  mytest                 myapp-deploy2-dc8f96497-nnkqn                1/1     Running   0          62m   10.244.3.194   k8s-node01   <none>           <none>  mytest                 myapp-deploy2-dc8f96497-w47dt                1/1     Running   0          62m   10.244.2.137   k8s-node02   <none>           <none>

myns 名称空间的pod，访问 mytest 名称空间的Service：myapp-clusterip2

[root@k8s-master cross_ns]# kubectl exec -it -n myns myapp-deploy1-5b9d78576c-wfw4n sh  / # cd /root/  ### 如下说明在同一名称空间下，通信无问题  ~ # ping myapp-clusterip1  PING myapp-clusterip1 (10.100.61.11): 56 data bytes  64 bytes from 10.100.61.11: seq=0 ttl=64 time=0.057 ms  64 bytes from 10.100.61.11: seq=1 ttl=64 time=0.071 ms  ………………  ~ #  ~ # wget myapp-clusterip1 -O myns.html  Connecting to myapp-clusterip1 (10.100.61.11:80)  myns.html            100%  ~ #  ~ # cat myns.html  Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>    ### 如下说明通过Service externalname类型，实现了Pod跨namespace名称空间与Service访问  ~ # ping myapp-clusterip1-externalname  PING myapp-clusterip1-externalname (10.100.201.103): 56 data bytes  64 bytes from 10.100.201.103: seq=0 ttl=64 time=0.050 ms  64 bytes from 10.100.201.103: seq=1 ttl=64 time=0.311 ms  ………………  ~ #  ~ # wget myapp-clusterip1-externalname -O mytest.html  Connecting to myapp-clusterip1-externalname (10.100.201.103:80)  mytest.html          100%  ~ #  ~ # cat mytest.html  Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>

mytest 名称空间的Pod，访问 myns 名称空间的Service：myapp-clusterip1

[root@k8s-master cross_ns]# kubectl exec -it -n mytest myapp-deploy2-dc8f96497-w47dt sh  / # cd /root/  ### 如下说明在同一名称空间下，通信无问题  ~ # ping myapp-clusterip2  PING myapp-clusterip2 (10.100.201.103): 56 data bytes  64 bytes from 10.100.201.103: seq=0 ttl=64 time=0.087 ms  64 bytes from 10.100.201.103: seq=1 ttl=64 time=0.073 ms  ………………  ~ #  ~ # wget myapp-clusterip2 -O mytest.html  Connecting to myapp-clusterip2 (10.100.201.103:80)  mytest.html          100%  ~ #  ~ # cat mytest.html  Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>    ### 如下说明通过Service externalname类型，实现了Pod跨namespace名称空间与Service访问  ~ # ping myapp-clusterip2-externalname  PING myapp-clusterip2-externalname (10.100.61.11): 56 data bytes  64 bytes from 10.100.61.11: seq=0 ttl=64 time=0.089 ms  64 bytes from 10.100.61.11: seq=1 ttl=64 time=0.071 ms  ………………  ~ #  ~ # wget myapp-clusterip2-externalname -O myns.html  Connecting to myapp-clusterip2-externalname (10.100.61.11:80)  myns.html            100%  ~ #  ~ # cat myns.html  Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>

由上可见，实现了Pod跨namespace名称空间与Service访问。

完毕！

原文出处：zhangblog -> http://www.zhangblog.com/2020/09/13/kubernetes14/