I'm trying to use NodePort with kind but somehow it doesn't want to work.

我正试着用Kind来使用NodePort，但不知何故它不想工作。

I've successfully deployed the following cluster:

我已经成功部署了以下集群：

apiVersion: kind.x-k8s.io/v1alpha4
kind: Cluster
nodes:
- role: control-plane
  extraPortMappings:
  - containerPort: 80
    hostPort: 30000
    listenAddress: "0.0.0.0" # Optional, defaults to "0.0.0.0"
    protocol: tcp # Optional, defaults to tcp
- role: worker

and then a very simple deployment:

然后是一个非常简单的部署：

apiVersion: apps/v1
kind: Deployment
metadata:
  name: hostname-deployment
  labels:
    app: hostname
spec:
  replicas: 2
  selector:
    matchLabels:
      app: hostname
  template:
    metadata:
      labels:
        app: hostname
    spec:
      containers:
      - name: hostname
        image: hostname:0.1
        ports:
        - containerPort: 80

and a service:

和一项服务：

apiVersion: v1
kind: Service
metadata:
  name: hostname-service
spec:
  type: NodePort
  selector:
    app: hostname
  ports:
    - protocol: TCP
      port: 80
      targetPort: 80
      nodePort: 30000

and I can connect to the service via e.g.

并且我可以通过例如

(in one terminal)
k port-forward service/hostname-service 8080:80
Forwarding from 127.0.0.1:8080 -> 80

(another one)
curl localhost:8080
hostname: hostname-deployment-75c9fd6584-ddc59 at Wed, 17 Jun 2020 15:38:33 UTC

But I cannot connect to the service via the exposed NodePort

但我无法通过公开的NodePort连接到该服务

curl -v localhost:30000
* Rebuilt URL to: localhost:30000/
*   Trying 127.0.0.1...
* TCP_NODELAY set
* Connected to localhost (127.0.0.1) port 30000 (#0)
> GET / HTTP/1.1
> Host: localhost:30000
> User-Agent: curl/7.58.0
> Accept: */*
>
* Recv failure: Connection reset by peer
* stopped the pause stream!
* Closing connection 0
curl: (56) Recv failure: Connection reset by peer

kubectl get all output:

Kubectl获取所有输出：

NAME                                       READY   STATUS    RESTARTS   AGE
pod/hostname-deployment-75c9fd6584-ddc59   1/1     Running   0          34m
pod/hostname-deployment-75c9fd6584-tg8db   1/1     Running   0          34m

NAME                       TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)        AGE
service/hostname-service   NodePort    10.107.104.231   <none>        80:30000/TCP   34m
service/kubernetes         ClusterIP   10.96.0.1        <none>        443/TCP        35m

NAME                                  READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/hostname-deployment   2/2     2            2           34m

NAME                                             DESIRED   CURRENT   READY   AGE
replicaset.apps/hostname-deployment-75c9fd6584   2         2         2       34m

Kind cluster configuration needs to be like below

种类群集配置需要如下所示

apiVersion: kind.x-k8s.io/v1alpha4
kind: Cluster
nodes:
- role: control-plane
  extraPortMappings:
  - containerPort: 30000
    hostPort: 30000
    listenAddress: "0.0.0.0" # Optional, defaults to "0.0.0.0"
    protocol: tcp # Optional, defaults to tcp
- role: worker

This file is then passed to your creation command as kind create cluster --config=config.yaml (according to docs).

然后，该文件以KIND CREATE CLUSTER--CONFIG=config.yaml的形式传递给您的创建命令(根据文档)。

Actually doing what Arghya Sadhu suggested worked. Not sure why the answer got deleted.

事实上，按照阿迦撒督的建议去做是有效的。不知道为什么答案被删除了。

apiVersion: kind.x-k8s.io/v1alpha4
kind: Cluster
nodes:
- role: control-plane
  extraPortMappings:
  - containerPort: 30000
    hostPort: 30000
    listenAddress: "0.0.0.0"
    protocol: tcp
- role: worker

I will share my answer as i tried out this today by using multi node kind cluster and another important thing is that needs to take care about kubectl client and server versions. I believe this will help someone.

我将分享我的答案，因为我今天尝试使用多节点类型的集群，另一件重要的事情是需要注意kubectl客户端和服务器版本。我相信这会对某些人有所帮助。

Step 1

步骤1

kind-api-cluster.yml

Kind-api-cluster.yml

kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
  extraPortMappings:
  - containerPort: 30950
    hostPort: 30950
    listenAddress: "127.0.0.1"
    protocol: TCP
- role: worker
- role: worker

Step 2
Kubernetes service and deployment file. ( If you note here, the kind's control-plane hostPort and kubernetes service's nodePort should be the same )

步骤2：Kubernetes服务和部署文件。(请注意，该类的控制平面HostPort和Kubernetes服务的nodePort应该是相同的)

api.yml

api.yml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: api
  labels:
    app: api
spec:
  replicas: 3
  selector:
    matchLabels:
      app: api
  template:
    metadata:
      labels:
        app: api
    spec:
      containers:
        - name: api
          image: api
          imagePullPolicy: Never
          ports:
            - containerPort: 8080
---
apiVersion: v1
kind: Service
metadata:
  name: api-svc
spec:
  selector:
    app: api
  type: NodePort
  ports:
    - name: http
      nodePort: 30950
      targetPort: 8080
      port: 8080

Additional Notes:-

其他备注：

kubectl version

Kubectl版本

Client Version: version.Info{Major:"1", Minor:"20+", GitVersion:"v1.20.4-dirty", GitCommit:"e87da0bd6e03ec3fea7933c4b5263d151aafd07c", GitTreeState:"dirty", BuildDate:"2021-03-15T09:55:27Z", GoVersion:"go1.16.2", Compiler:"gc", Platform:"darwin/amd64"}
Server Version: version.Info{Major:"1", Minor:"20", GitVersion:"v1.20.2", GitCommit:"faecb196815e248d3ecfb03c680a4507229c2a56", GitTreeState:"clean", BuildDate:"2021-01-21T01:11:42Z", GoVersion:"go1.15.5", Compiler:"gc", Platform:"linux/amd64"}

kind --version

Kind--版本

kind version 0.10.0

You shouldn't use nodeport with local IP.
Instead you should:

本端IP不能使用nodeport。相反，您应该：

1. use command : kubectl get nodes -o wide


2. note down the IP of the node


3. curl :30000

or

或

2. docker exec -it /bin/bash

   
   



3. curl localhost:30000

   
   

in the kind config yaml

在KIND配置YAML中

containerPort must be the same with the 'NodePort' in service config.
in your case is 30000

容器端口必须与服务配置中的‘NodePort’相同。在你的情况下是30000

tips: hostPort can be any port ,it does not need to be the same with containerPort

小贴士：主机端口可以是任何端口，它不需要与容器端口相同

kubectl port-forward svc/hostname-service 30000:80

Now your service will be accessed on host port 30000

现在可以通过主机端口30000访问您的服务

I would try removing the nodePort: 30000 and check that the newly assigned nodePort works with the loopback interface. If you do an update of the service, kube has a problem with satically asigned nodePorts, you have to manige yourself the port collision. Trying to delete the deployment and redeploy could also help.

我会尝试删除nodePort：30000，并检查新分配的nodePort是否与环回接口一起工作。如果您进行服务更新，Kube在静态分配的nodePorts方面有问题，您必须管理自己的端口冲突。尝试删除部署并重新部署也可能有所帮助。

I got it worked by two options:

我通过两种选择让它发挥作用：

1. Either you keep the same nodePort config while creating the service and the one used to create the kind cluster.


2. The second is to keep any nodePort in the service manifest just use the port forward command

kubectl port-forward svc/hostname-service [assigned node port to service]:80

Kubectl port-转发服务/主机名-服务[分配给服务的节点端口]：80

Then access it using localhost:[nodePort]

然后使用本地主机访问它：[nodePort]