在k8s集群上用Deployment运行一个无状态的应用程序nginx

2018-05-26 3419点热度 2人点赞 0条评论

前言

既然k8s集群机已经搭建好，那就要试用一下，在dashboard的工作负载中没有任何东西，点击现在就部署容器应用，进入官方文档链接，就是本文所要做的事情。

目标

创建一个nginx deployment
用kubectl列出此deployment的一些信息
更新deployment
把程序发布出去，在宿主机能够访问

开始之前

在开始之前，已经搭建了一个k8s集群，搭建集群链接在fedora server 28上基于Kubeadm搭建Kubernetes v1.10.3全记录，并且在命令行中kubectl可以与集群进行交互。
k8s服务器版本必须是v1.9或者更高，可以使用 kubectl version -o yaml 查看版本。

[wangxianfeng@k8s-m1 ~]$ kubectl version -o yaml
clientVersion:
  buildDate: 2018-05-21T09:17:39Z
  compiler: gc
  gitCommit: 2bba0127d85d5a46ab4b778548be28623b32d0b0
  gitTreeState: clean
  gitVersion: v1.10.3
  goVersion: go1.9.3
  major: "1"
  minor: "10"
  platform: linux/amd64
serverVersion:
  buildDate: 2018-05-21T09:05:37Z
  compiler: gc
  gitCommit: 2bba0127d85d5a46ab4b778548be28623b32d0b0
  gitTreeState: clean
  gitVersion: v1.10.3
  goVersion: go1.9.3
  major: "1"
  minor: "10"
  platform: linux/amd64

创建并开放一个nginx deployment

可以用k8s Deployment运行一个应用程序，可以用YAML文件定义一个Deployment，以下文件用一个YAML文件定义了一个nginx docker镜像：

apiVersion: apps/v1 # for versions before 1.9.0 use apps/v1beta2
kind: Deployment
metadata:
  name: nginx-deployment
spec:
  selector:
    matchLabels:
      app: nginx
  replicas: 2 # tells deployment to run 2 pods matching the template
  template: # create pods using pod definition in this template
    metadata:
      # unlike pod-nginx.yaml, the name is not included in the meta data as a unique name is
      # generated from the deployment name
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        imagePullPolicy: IfNotPresent #或者使用Never，默认Always
        ports:
        - containerPort: 80

保存文件名为deployment.yaml

用yaml文件创建一个Deployment

$ kubectl apply -f deployment.yaml
deployment.apps "nginx-deployment" created

列出此Deployment的一些信息

kubectl describe deployment nginx-deployment

输出如下：

[wangxianfeng@k8s-m1 k8s]$ kubectl describe deployment nginx-deployment
Name:                   nginx-deployment
Namespace:              default
CreationTimestamp:      Sat, 26 May 2018 04:34:25 +0800
Labels:                 <none>
Annotations:            deployment.kubernetes.io/revision=1
                        kubectl.kubernetes.io/last-applied-configuration={"apiVersion":"apps/v1","kind":"Deployment","metadata":{"annotations":{},"name":"nginx-deployment","namespace":"default"},"spec":{"replicas":2,"selecto...
Selector:               app=nginx
Replicas:               2 desired | 2 updated | 2 total | 0 available | 2 unavailable
StrategyType:           RollingUpdate
MinReadySeconds:        0
RollingUpdateStrategy:  25% max unavailable, 25% max surge
Pod Template:
  Labels:  app=nginx
  Containers:
   nginx:
    Image:        nginx
    Port:         80/TCP
    Host Port:    0/TCP
    Environment:  <none>
    Mounts:       <none>
  Volumes:        <none>
Conditions:
  Type           Status  Reason
  ----           ------  ------
  Available      False   MinimumReplicasUnavailable
  Progressing    True    ReplicaSetUpdated
OldReplicaSets:  <none>
NewReplicaSet:   nginx-deployment-666865b5dd (2/2 replicas created)
Events:
  Type    Reason             Age   From                   Message
  ----    ------             ----  ----                   -------
  Normal  ScalingReplicaSet  1m    deployment-controller  Scaled up replica set nginx-deployment-666865b5dd to 2

列出此deployment创建的pods

kubectl get pods -l app=nginx

输出如下：

[wangxianfeng@k8s-m1 k8s]$ kubectl get pods -l app=nginx
NAME                                READY     STATUS    RESTARTS   AGE
nginx-deployment-666865b5dd-89d45   0/1       Pending   0          4m
nginx-deployment-666865b5dd-kwl5t   0/1       Pending   0          4m

pod的状态为Pending

列出pod的信息

kubectl describe pod nginx-deployment-666865b5dd-89d45

nginx-deployment-666865b5dd-89d45 为上边列出的pod中的一个。
输出如下：

[wangxianfeng@k8s-m1 k8s]$ kubectl describe pod nginx-deployment-666865b5dd-89d45
Name:           nginx-deployment-666865b5dd-89d45
Namespace:      default
Node:           <none>
Labels:         app=nginx
                pod-template-hash=2224216188
Annotations:    <none>
Status:         Pending
IP:             
Controlled By:  ReplicaSet/nginx-deployment-666865b5dd
Containers:
  nginx:
    Image:        nginx
    Port:         80/TCP
    Host Port:    0/TCP
    Environment:  <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-dqzwh (ro)
Conditions:
  Type           Status
  PodScheduled   False 
Volumes:
  default-token-dqzwh:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-dqzwh
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason            Age                From               Message
  ----     ------            ----               ----               -------
  Warning  FailedScheduling  36s (x26 over 6m)  default-scheduler  0/2 nodes are available: 2 node(s) had taints that the pod didn't tolerate.

得到错误信息如下：

0/2 nodes are available: 2 node(s) had taints that the pod didn't tolerate.

搜索此错误信息，也有人问过这个问题：
Kubeadm and the Risks of Scheduling Pods on Master Node (Pods always Pending)

运行以下命令可以解决：

kubectl taint nodes --all node-role.kubernetes.io/master-

但是，带来了一个问题，就是在master节点上运行节工作负载。先不考虑，那么多了，先运行起来再说吧。

[wangxianfeng@k8s-m1 k8s]$ kubectl get pods -l app=nginx
NAME                                READY     STATUS    RESTARTS   AGE
nginx-deployment-666865b5dd-89d45   1/1       Running   0          29m
nginx-deployment-666865b5dd-kwl5t   1/1       Running   0          29m

现在状态全部都是running了。对比以下pod的输出(pod的2个副本一个在k8s-m1上，一个在k8s-n1上)：
k8s-m1上的pod信息：

[wangxianfeng@k8s-m1 k8s]$ kubectl describe pod nginx-deployment-666865b5dd-kwl5t
Name:           nginx-deployment-666865b5dd-kwl5t
Namespace:      default
Node:           k8s-m1/192.168.2.186
Start Time:     Sat, 26 May 2018 04:52:55 +0800
Labels:         app=nginx
                pod-template-hash=2224216188
Annotations:    <none>
Status:         Running
IP:             10.244.0.4
Controlled By:  ReplicaSet/nginx-deployment-666865b5dd
Containers:
  nginx:
    Container ID:   docker://9aa9847e92dbeea296e5ba7872a391325d1c4800ff1ae0e426b993686ea67695
    Image:          nginx
    Image ID:       docker-pullable://nginx@sha256:0fb320e2a1b1620b4905facb3447e3d84ad36da0b2c8aa8fe3a5a81d1187b884
    Port:           80/TCP
    Host Port:      0/TCP
    State:          Running
      Started:      Sat, 26 May 2018 04:53:37 +0800
    Ready:          True
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-dqzwh (ro)
Conditions:
  Type           Status
  Initialized    True 
  Ready          True 
  PodScheduled   True 
Volumes:
  default-token-dqzwh:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-dqzwh
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason                 Age                 From               Message
  ----     ------                 ----                ----               -------
  Warning  FailedScheduling       15m (x57 over 30m)  default-scheduler  0/2 nodes are available: 2 node(s) had taints that the pod didn't tolerate.
  Normal   SuccessfulMountVolume  12m                 kubelet, k8s-m1    MountVolume.SetUp succeeded for volume "default-token-dqzwh"
  Normal   Pulling                12m                 kubelet, k8s-m1    pulling image "nginx"
  Normal   Pulled                 11m                 kubelet, k8s-m1    Successfully pulled image "nginx"
  Normal   Created                11m                 kubelet, k8s-m1    Created container
  Normal   Started                11m                 kubelet, k8s-m1    Started container

k8s-n1上的pod信息：

[wangxianfeng@k8s-m1 k8s]$ kubectl describe pod nginx-deployment-666865b5dd-89d45
Name:           nginx-deployment-666865b5dd-89d45
Namespace:      default
Node:           k8s-n1/192.168.2.189
Start Time:     Sat, 26 May 2018 04:52:57 +0800
Labels:         app=nginx
                pod-template-hash=2224216188
Annotations:    <none>
Status:         Running
IP:             10.244.1.4
Controlled By:  ReplicaSet/nginx-deployment-666865b5dd
Containers:
  nginx:
    Container ID:   docker://504e8bc1a6231655644c497e7a132fb6465d916fe459b5937ee53c800af54be8
    Image:          nginx
    Image ID:       docker-pullable://nginx@sha256:0fb320e2a1b1620b4905facb3447e3d84ad36da0b2c8aa8fe3a5a81d1187b884
    Port:           80/TCP
    Host Port:      0/TCP
    State:          Running
      Started:      Sat, 26 May 2018 04:53:35 +0800
    Ready:          True
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-dqzwh (ro)
Conditions:
  Type           Status
  Initialized    True 
  Ready          True 
  PodScheduled   True 
Volumes:
  default-token-dqzwh:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-dqzwh
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason                 Age                 From               Message
  ----     ------                 ----                ----               -------
  Warning  FailedScheduling       18m (x57 over 33m)  default-scheduler  0/2 nodes are available: 2 node(s) had taints that the pod didn't tolerate.
  Normal   SuccessfulMountVolume  15m                 kubelet, k8s-n1    MountVolume.SetUp succeeded for volume "default-token-dqzwh"
  Normal   Pulling                15m                 kubelet, k8s-n1    pulling image "nginx"
  Normal   Pulled                 14m                 kubelet, k8s-n1    Successfully pulled image "nginx"
  Normal   Created                14m                 kubelet, k8s-n1    Created container
  Normal   Started                14m                 kubelet, k8s-n1    Started container

更新deployment

可以用一个新的YAML文件更新一个deployment，复制deployment.yaml文件为deployment_update.yaml，编辑deployment_update.yaml文件，更新nginx镜像版本为1.8

kind: Deployment
metadata:
  name: nginx-deployment
spec:
  selector:
    matchLabels:
      app: nginx
  replicas: 2 # tells deployment to run 2 pods matching the template
  template: # create pods using pod definition in this template
    metadata:
      # unlike pod-nginx.yaml, the name is not included in the meta data as a unique name is
      # generated from the deployment name
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.8 # Update the version of nginx to 1.8
        ports:
        - containerPort: 80

应用更新

[wangxianfeng@k8s-m1 k8s]$ kubectl apply -f deployment_update.yaml
deployment.apps "nginx-deployment" configured

查看deployment新创建的pods信息

kubectl get pods -l app=nginx

输出如下：

[wangxianfeng@k8s-m1 k8s]$ kubectl get pods -l app=nginx
NAME                               READY     STATUS    RESTARTS   AGE
nginx-deployment-cf7f48d4c-dchd7   1/1       Running   0          1m
nginx-deployment-cf7f48d4c-m4hg4   1/1       Running   0          1m

查看一个详细的信息：

[wangxianfeng@k8s-m1 k8s]$ kubectl describe pod nginx-deployment-cf7f48d4c-dchd
Name:           nginx-deployment-cf7f48d4c-dchd7
Namespace:      default
Node:           k8s-n1/192.168.2.189
Start Time:     Sat, 26 May 2018 05:17:35 +0800
Labels:         app=nginx
                pod-template-hash=793904807
Annotations:    <none>
Status:         Running
IP:             10.244.1.5
Controlled By:  ReplicaSet/nginx-deployment-cf7f48d4c
Containers:
  nginx:
    Container ID:   docker://79e4a3ec93024455130ceaeb8babd3e6be3221f4a044f8f1b7695a59d279fd60
    Image:          nginx:1.8
    Image ID:       docker-pullable://nginx@sha256:c97ee70c4048fe79765f7c2ec0931957c2898f47400128f4f3640d0ae5d60d10
    Port:           80/TCP
    Host Port:      0/TCP
    State:          Running
      Started:      Sat, 26 May 2018 05:18:21 +0800
    Ready:          True
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-dqzwh (ro)
Conditions:
  Type           Status
  Initialized    True 
  Ready          True 
  PodScheduled   True 
Volumes:
  default-token-dqzwh:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-dqzwh
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type    Reason                 Age   From               Message
  ----    ------                 ----  ----               -------
  Normal  Scheduled              1m    default-scheduler  Successfully assigned nginx-deployment-cf7f48d4c-dchd7 to k8s-n1
  Normal  SuccessfulMountVolume  1m    kubelet, k8s-n1    MountVolume.SetUp succeeded for volume "default-token-dqzwh"
  Normal  Pulling                1m    kubelet, k8s-n1    pulling image "nginx:1.8"
  Normal  Pulled                 1m    kubelet, k8s-n1    Successfully pulled image "nginx:1.8"
  Normal  Created                1m    kubelet, k8s-n1    Created container
  Normal  Started                1m    kubelet, k8s-n1    Started container

可以看到nginx容器镜像已经变成了1.8版本。

增加副本数量

可以在YAML文件中简单修改副本的数量来扩展Deployment的pod。复制deployment.yaml文件为deployment_scale.yaml，编辑deployment_scale.yaml文件，把replicas修改为4：

kind: Deployment
metadata:
  name: nginx-deployment
spec:
  selector:
    matchLabels:
      app: nginx
  replicas: 4 # tells deployment to run 4 pods matching the template
  template: # create pods using pod definition in this template
    metadata:
      # unlike pod-nginx.yaml, the name is not included in the meta data as a unique name is
      # generated from the deployment name
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx # nginx版本还修改为latest，我比较喜欢用新版本
        ports:
        - containerPort: 80

应用更新，查看pods信息(中间过程也都贴了，期间有pod创建，有pod销毁，最后有4个状态为running）：

[wangxianfeng@k8s-m1 k8s]$ kubectl apply -f deployment_scale.yaml
deployment.apps "nginx-deployment" configured
[wangxianfeng@k8s-m1 k8s]$ kubectl get pods -l app=nginx
NAME                                READY     STATUS              RESTARTS   AGE
nginx-deployment-666865b5dd-h4d4l   0/1       ContainerCreating   0          11s
nginx-deployment-666865b5dd-wmlmk   0/1       ContainerCreating   0          10s
nginx-deployment-cf7f48d4c-dchd7    1/1       Running             0          8m
nginx-deployment-cf7f48d4c-m4hg4    1/1       Running             0          9m
nginx-deployment-cf7f48d4c-mv5vf    0/1       ContainerCreating   0          12s
[wangxianfeng@k8s-m1 k8s]$ kubectl get pods -l app=nginx
NAME                                READY     STATUS              RESTARTS   AGE
nginx-deployment-666865b5dd-h4d4l   1/1       Running             0          30s
nginx-deployment-666865b5dd-wmlmk   1/1       Running             0          29s
nginx-deployment-666865b5dd-xwnds   0/1       ContainerCreating   0          11s
nginx-deployment-666865b5dd-zntfs   0/1       ContainerCreating   0          8s
nginx-deployment-cf7f48d4c-dchd7    0/1       Terminating         0          8m
nginx-deployment-cf7f48d4c-m4hg4    1/1       Running             0          9m
[wangxianfeng@k8s-m1 k8s]$ kubectl get pods -l app=nginx
NAME                                READY     STATUS              RESTARTS   AGE
nginx-deployment-666865b5dd-h4d4l   1/1       Running             0          33s
nginx-deployment-666865b5dd-wmlmk   1/1       Running             0          32s
nginx-deployment-666865b5dd-xwnds   0/1       ContainerCreating   0          14s
nginx-deployment-666865b5dd-zntfs   0/1       ContainerCreating   0          11s
nginx-deployment-cf7f48d4c-dchd7    0/1       Terminating         0          8m
nginx-deployment-cf7f48d4c-m4hg4    1/1       Running             0          9m
[wangxianfeng@k8s-m1 k8s]$ kubectl get pods -l app=nginx
NAME                                READY     STATUS    RESTARTS   AGE
nginx-deployment-666865b5dd-h4d4l   1/1       Running   0          50s
nginx-deployment-666865b5dd-wmlmk   1/1       Running   0          49s
nginx-deployment-666865b5dd-xwnds   1/1       Running   0          31s
nginx-deployment-666865b5dd-zntfs   1/1       Running   0          28s

删除一个Deployment

根据deployment的名字进行删除：

kubectl delete deployment nginx-deployment

查看pod运行在哪个节点上

kubectl get pods -l app=nginx -o wide

根据lable进行筛选，-o 表示输出格式，wide表示多输出一些信息，如IP，pod运行在哪个节点上等等。其他输出格式可以选择yaml，json等。具体参考Overview of kubectl

查看pod的IP信息

kubectl get pods -l app=nginx -o yaml|grep podIP

用service连接应用

官方文档链接Connecting Applications with Services

创建一个Service

kubectl expose deployment/nginx-deployment

此语句根据名称为nginx-deployment的deployment创建了一个service，也可以使用yaml文件创建service，并进行更多的定制化，等价于以下yaml文件：

apiVersion: v1
kind: Service
metadata:
  name: nginx-deployment
  labels:
    app: nginx
spec:
  ports:
  - port: 80
    protocol: TCP
  selector:
    app: nginx

使用 kubectl create -f 进行创建。

创建完成之后，通过以下命令查看是否创建成功：

kubectl get svc nginx-deployment

输出类似如下：

[wangxianfeng@k8s-m1 k8s]$ kubectl get svc nginx-deployment
NAME               TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)   AGE
nginx-deployment   ClusterIP   10.102.140.205   <none>        80/TCP    10m

可以看到该服务监听在80端口，集群IP是10.102.140.205。

可以通过describe命令查看service的pod详情：

kubectl describe svc nginx-deployment

输出类似如下：

[wangxianfeng@k8s-m1 k8s]$ kubectl describe svc nginx-deployment
Name:              nginx-deployment
Namespace:         default
Labels:            app=nginx
Annotations:       <none>
Selector:          app=nginx
Type:              ClusterIP
IP:                10.102.140.205
Port:              <unset>  80/TCP
TargetPort:        80/TCP
Endpoints:         10.244.0.6:80,10.244.0.8:80,10.244.0.9:80 + 1 more...
Session Affinity:  None
Events:            <none>

访问Service

Kubernetes支持2中主要的模型用于查找服务：环境变量和DNS。

环境变量

pod运行的时候kubelet会添加一些环境变量，可以用以下命令查看pod的环境变量：

[wangxianfeng@k8s-m1 k8s]$ kubectl get pod
NAME                                READY     STATUS    RESTARTS   AGE
nginx-deployment-666865b5dd-h4d4l   1/1       Running   0          1h
nginx-deployment-666865b5dd-wmlmk   1/1       Running   0          1h
nginx-deployment-666865b5dd-xwnds   1/1       Running   0          1h
nginx-deployment-666865b5dd-zntfs   1/1       Running   0          1h
[wangxianfeng@k8s-m1 k8s]$ kubectl exec nginx-deployment-666865b5dd-h4d4l -- printenv | grep SERVICE
KUBERNETES_SERVICE_HOST=10.96.0.1
KUBERNETES_SERVICE_PORT=443
KUBERNETES_SERVICE_PORT_HTTPS=443

环境变量中并没有Service的相关信息，因为创建pod在创建Service之前。先把nginx-deployment的replicas改成0，然后再改成4，再看看pod的节点分布情况：

kubectl scale deployment nginx-deployment --replicas=0; kubectl scale deployment nginx-deployment --replicas=4;

这些分布比较均匀了：

[wangxianfeng@k8s-m1 k8s]$ kubectl get pods -l app=nginx -o wide

NAME                                READY     STATUS    RESTARTS   AGE       IP            NODE
nginx-deployment-666865b5dd-kb9gc   1/1       Running   0          57s       10.244.1.7    k8s-n1
nginx-deployment-666865b5dd-m77lv   1/1       Running   0          57s       10.244.0.12   k8s-m1
nginx-deployment-666865b5dd-m9kcd   1/1       Running   0          57s       10.244.0.14   k8s-m1
nginx-deployment-666865b5dd-z6jld   1/1       Running   0          57s       10.244.0.13   k8s-m1

再次查看环境变量，已经有了Service相关信息：

[wangxianfeng@k8s-m1 k8s]$ kubectl exec nginx-deployment-666865b5dd-kb9gc -- printenv | grep SERVICE
KUBERNETES_SERVICE_PORT_HTTPS=443
NGINX_DEPLOYMENT_SERVICE_HOST=10.102.140.205
KUBERNETES_SERVICE_HOST=10.96.0.1
NGINX_DEPLOYMENT_SERVICE_PORT=80
KUBERNETES_SERVICE_PORT=443

DNS

使用如下命令查看是否已经在运行dns服务

kubectl get services kube-dns --namespace=kube-system

NAME       TYPE        CLUSTER-IP   EXTERNAL-IP   PORT(S)         AGE
kube-dns   ClusterIP   10.96.0.10   <none>        53/UDP,53/TCP   1d

运行一个curl的docker容器进行测试

kubectl run curl --image=radial/busyboxplus:curl -i --tty

[wangxianfeng@k8s-m1 k8s]$ kubectl run curl --image=radial/busyboxplus:curl -i --tty
If you don't see a command prompt, try pressing enter.
[ root@curl-775f9567b5-knshg:/ ]$ nslookup nginx-deployment
Server:    10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local

nslookup: can't resolve 'nginx-deployment'

从宿主机访问Service（Exposing the Service）

两种方式：NodePorts and LoadBalancers
用以下命令查看是否已经开放了nodeport及是否具有外部IP

kubectl get svc nginx-deployment -o yaml | grep nodePort -C 5
kubectl get nodes -o yaml | grep ExternalIP -C 1

现在编辑Service，添加nodePort。首先删除nginx-deployment

kubectl delete svc nginx-deployment

创建文件nginx_service.yaml：

apiVersion: v1
kind: Service
metadata:
  name: nginx-deployment
  labels:
    app: nginx
spec:
  type: NodePort
  ports:
  - port: 8080
    protocol: TCP
    targetPort: 80
  selector:
    app: nginx

使用如下命令创建Service：

kubectl create -f nginx_service.yaml

再看看是不是已经有了nodeport：

[wangxianfeng@k8s-m1 k8s]$ kubectl get svc nginx-deployment -o yaml | grep nodePort -C 5
  uid: 62db4297-6078-11e8-a407-0211322b88e3
spec:
  clusterIP: 10.109.119.99
  externalTrafficPolicy: Cluster
  ports:
  - nodePort: 32494
    port: 8080
    protocol: TCP
    targetPort: 80
  selector:
    app: nginx

现在访问宿主机的32494端口即可访问Service了：

curl http://192.168.2.186:32494

浏览器里边看一把

现在把NodePort改成LoadBalancer试试。

[wangxianfeng@k8s-m1 k8s]$ kubectl get svc nginx-deployment -o yaml | grep nodePort -C 5
  uid: 85b913ad-6079-11e8-a407-0211322b88e3
spec:
  clusterIP: 10.111.180.9
  externalTrafficPolicy: Cluster
  ports:
  - nodePort: 32123
    port: 8080
    protocol: TCP
    targetPort: 80
  selector:
    app: nginx

差不多，不知道啥区别，不管了。

本作品采用知识共享署名 4.0 国际许可协议进行许可