amazon-ec2 - kubernetes 入口 aws 部署负载均衡器待处理

Question

简而言之，这些是我已经完成的步骤:

上线 2 新 t3 - small aws 中的实例，用 key 预先标记kubernetes.io/cluster/<cluster-name>和值(value)member .

用相同的标签标记新的安全并打开所有提到的端口
这里 -
https://kubernetes.io/docs/setup/independent/install-kubeadm/#check-required-ports

已更改 hostname到 curl
http://169.254.169.254/latest/meta-data/local-hostname 的输出并经过验证
与 hostnamectl

重启

配置awshttps://docs.aws.amazon.com/cli/latest/userguide/cli-chap-configure.html

已创建 IAM role具有完整 ( "*" ) 权限并分配给 EC2
实例。

已安装 kubelet kubeadm kubectl使用 apt-get

已创建 /etc/default/kubelet有内容KUBELET_EXTRA_ARGS=--cloud-provider=aws

冉 kubeadm init --pod-network-cidr=10.244.0.0/16在一个实例上
并使用输出到 kubeadm join ...其他节点。

已安装 Helm .

已安装 ingress controller使用默认后端。

以前我已经尝试过上述步骤，但是，根据 kubernetes.github.io 上的说明安装了入口。 .两者都以相同的状态结束， EXTERNAL-IP如 <pending> .

当前状态是:
kubectl get pods --all-namespaces -o wide

NAMESPACE     NAME                                                                   IP              NODE                                           
ingress       ingress-nginx-ingress-controller-77d989fb4d-qz4f5                      10.244.1.13     ip-YYY-YY-Y-YYY.ap-south-1.compute.internal               
ingress       ingress-nginx-ingress-default-backend-7f7bf55777-dhj75                 10.244.1.12     ip-YYY-YY-Y-YYY.ap-south-1.compute.internal               
kube-system   coredns-86c58d9df4-bklt8                                               10.244.1.14     ip-YYY-YY-Y-YYY.ap-south-1.compute.internal               
kube-system   coredns-86c58d9df4-ftn8q                                               10.244.1.16     ip-YYY-YY-Y-YYY.ap-south-1.compute.internal               
kube-system   etcd-ip-XXX-XX-XX-XXX.ap-south-1.compute.internal                      172.31.12.119   ip-XXX-XX-XX-XXX.ap-south-1.compute.internal              
kube-system   kube-apiserver-ip-XXX-XX-XX-XXX.ap-south-1.compute.internal            172.31.12.119   ip-XXX-XX-XX-XXX.ap-south-1.compute.internal              
kube-system   kube-controller-manager-ip-XXX-XX-XX-XXX.ap-south-1.compute.internal   172.31.12.119   ip-XXX-XX-XX-XXX.ap-south-1.compute.internal              
kube-system   kube-flannel-ds-amd64-87k8p                                            172.31.12.119   ip-XXX-XX-XX-XXX.ap-south-1.compute.internal              
kube-system   kube-flannel-ds-amd64-f4wft                                            172.31.3.106    ip-YYY-YY-Y-YYY.ap-south-1.compute.internal               
kube-system   kube-proxy-79cp2                                                       172.31.3.106    ip-YYY-YY-Y-YYY.ap-south-1.compute.internal               
kube-system   kube-proxy-sv7md                                                       172.31.12.119   ip-XXX-XX-XX-XXX.ap-south-1.compute.internal              
kube-system   kube-scheduler-ip-XXX-XX-XX-XXX.ap-south-1.compute.internal            172.31.12.119   ip-XXX-XX-XX-XXX.ap-south-1.compute.internal              
kube-system   tiller-deploy-5b7c66d59c-fgwcp                                         10.244.1.15     ip-YYY-YY-Y-YYY.ap-south-1.compute.internal  

kubectl get svc --all-namespaces -o wide

NAMESPACE     NAME                                    TYPE           CLUSTER-IP       EXTERNAL-IP   PORT(S)                      AGE   SELECTOR
default       kubernetes                              ClusterIP      10.96.0.1        <none>        443/TCP                      73m   <none>
ingress       ingress-nginx-ingress-controller        LoadBalancer   10.97.167.197    <pending>     80:32722/TCP,443:30374/TCP   59m   app=nginx-ingress,component=controller,release=ingress
ingress       ingress-nginx-ingress-default-backend   ClusterIP      10.109.198.179   <none>        80/TCP                       59m   app=nginx-ingress,component=default-backend,release=ingress
kube-system   kube-dns                                ClusterIP      10.96.0.10       <none>        53/UDP,53/TCP                73m   k8s-app=kube-dns
kube-system   tiller-deploy                           ClusterIP      10.96.216.119    <none>        44134/TCP                    67m   app=helm,name=tiller

kubectl describe service -n ingress ingress-nginx-ingress-controller
Name:                     ingress-nginx-ingress-controller
Namespace:                ingress
Labels:                   app=nginx-ingress
                          chart=nginx-ingress-1.4.0
                          component=controller
                          heritage=Tiller
                          release=ingress
Annotations:              service.beta.kubernetes.io/aws-load-balancer-proxy-protocol: *
Selector:                 app=nginx-ingress,component=controller,release=ingress
Type:                     LoadBalancer
IP:                       10.104.55.18
Port:                     http  80/TCP
TargetPort:               http/TCP
NodePort:                 http  32318/TCP
Endpoints:                10.244.1.20:80
Port:                     https  443/TCP
TargetPort:               https/TCP
NodePort:                 https  32560/TCP
Endpoints:                10.244.1.20:443
Session Affinity:         None
External Traffic Policy:  Cluster
Events:                   <none>

附加的 IAM 角色内联策略

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": "*",
            "Resource": "*"
        }
    ]
}

kubectl 获取节点 -o 宽

NAME                                           STATUS   ROLES    AGE     VERSION   INTERNAL-IP     EXTERNAL-IP     OS-IMAGE             KERNEL-VERSION   CONTAINER-RUNTIME
ip-172-31-12-119.ap-south-1.compute.internal   Ready    master   6d19h   v1.13.4   172.31.12.119   XX.XXX.XXX.XX   Ubuntu 16.04.5 LTS   4.4.0-1077-aws   docker://18.6.3
ip-172-31-3-106.ap-south-1.compute.internal    Ready    <none>   6d19h   v1.13.4   172.31.3.106    XX.XXX.XX.XXX   Ubuntu 16.04.5 LTS   4.4.0-1077-aws   docker://18.6.3

有人可以指出我在这里遗漏了什么，因为互联网上到处都是 Classic ELB会自动部署吗？

Answer 1

对于 AWS ELB(经典类型)，您必须

明确指定 --cloud-provider=aws在 kube 服务 list 中
位于/etc/kubernetes/manifests在主节点上:
kube-controller-manager.yaml
kube-apiserver.yaml

重启服务:
sudo systemctl daemon-reloadsudo systemctl restart kubelet

与其他命令一起，根据需要在底部或顶部添加。结果应该类似于:

在 kube-controller-manager.yaml

spec:
  containers:
  - command:
    - kube-controller-manager
    - --cloud-provider=aws

在 kube-apiserver.yaml

spec:
  containers:
  - command:
    - kube-apiserver
    - --cloud-provider=aws