首页 > 解决方案 > 设置 Kubernetes - 无法从 Pod 访问 API

问题描述

我正在尝试在(Ubuntu 16)VM 上设置一个基本的 Kubernetes 集群。我刚刚关注了入门文档,并期望有一个工作集群,但不幸的是,没有这样的运气 - 似乎没有 pod 无法连接到 Kubenernetes API。因为我是 Kubernetes 的新手,所以我很难找到哪里出了问题。提供脚本:

apt-get update && apt-get upgrade -y
apt-get install -y apt-transport-https curl
curl -s https://packages.cloud.google.com/apt/doc/apt-key.gpg | apt-key add -
cat <<EOF >/etc/apt/sources.list.d/kubernetes.list
deb https://apt.kubernetes.io/ kubernetes-xenial main
EOF
apt-get update
apt-get install -y kubelet kubeadm kubectl docker.io
apt-mark hold kubelet kubeadm kubectl
swapoff -a
sysctl net.bridge.bridge-nf-call-iptables=1
kubeadm init
mkdir -p /home/ubuntu/.kube
cp -i /etc/kubernetes/admin.conf /home/ubuntu/.kube/config
chown -R ubuntu:ubuntu /home/ubuntu/.kube
runuser -l ubuntu -c "kubectl apply -f \"https://cloud.weave.works/k8s/net?k8s-version=$(kubectl version | base64 | tr -d '\n')\""
runuser -l ubuntu -c "kubectl taint nodes --all node-role.kubernetes.io/master-"

安装似乎很好。

ubuntu@packer-Ubuntu-16:~$ kubectl get pods -o wide --all-namespaces
NAMESPACE     NAME                                       READY   STATUS             RESTARTS   AGE   IP               NODE               NOMINATED NODE   READINESS GATES
kube-system   coredns-86c58d9df4-lbp46                   0/1     CrashLoopBackOff   7          18m   10.32.0.2        packer-ubuntu-16   <none>           <none>
kube-system   coredns-86c58d9df4-t8nnn                   0/1     CrashLoopBackOff   7          18m   10.32.0.3        packer-ubuntu-16   <none>           <none>
kube-system   etcd-packer-ubuntu-16                      1/1     Running            0          17m   145.100.100.100  packer-ubuntu-16   <none>           <none>
kube-system   kube-apiserver-packer-ubuntu-16            1/1     Running            0          18m   145.100.100.100  packer-ubuntu-16   <none>           <none>
kube-system   kube-controller-manager-packer-ubuntu-16   1/1     Running            0          17m   145.100.100.100  packer-ubuntu-16   <none>           <none>
kube-system   kube-proxy-dwhhf                           1/1     Running            0          18m   145.100.100.100  packer-ubuntu-16   <none>           <none>
kube-system   kube-scheduler-packer-ubuntu-16            1/1     Running            0          17m   145.100.100.100  packer-ubuntu-16   <none>           <none>
kube-system   weave-net-sfvz5                            2/2     Running            0          18m   145.100.100.100  packer-ubuntu-16   <none>           <none>

问题:Kubernetes pod 的 IP 是主机的 eth0 的 IP(145.100.100.100)是否正常?对我来说似乎很奇怪,我希望他们有一个虚拟 IP?

如您所见,corednspod 正在崩溃,因为它无法访问 API。

据我了解,服务:

ubuntu@packer-Ubuntu-16:~$ kubectl get svc
NAME         TYPE        CLUSTER-IP   EXTERNAL-IP   PORT(S)   AGE
kubernetes   ClusterIP   10.96.0.1    <none>        443/TCP   22m

CoreDNS 崩溃,因为 API 无法访问:

ubuntu@packer-Ubuntu-16:~$ kubectl logs -n kube-system coredns-86c58d9df4-lbp46
.:53
2018-12-06T12:54:28.481Z [INFO] CoreDNS-1.2.6
2018-12-06T12:54:28.481Z [INFO] linux/amd64, go1.11.2, 756749c
CoreDNS-1.2.6
linux/amd64, go1.11.2, 756749c
 [INFO] plugin/reload: Running configuration MD5 = f65c4821c8a9b7b5eb30fa4fbc167769
E1206 12:54:53.482269       1 reflector.go:205] github.com/coredns/coredns/plugin/kubernetes/controller.go:318: Failed to list *v1.Namespace: Get https://10.96.0.1:443/api/v1/namespaces?limit=500&resourceVersion=0: dial tcp 10.96.0.1:443: i/o timeout
E1206 12:54:53.482363       1 reflector.go:205] github.com/coredns/coredns/plugin/kubernetes/controller.go:311: Failed to list *v1.Service: Get https://10.96.0.1:443/api/v1/services?limit=500&resourceVersion=0: dial tcp 10.96.0.1:443: i/o timeout
E1206 12:54:53.482540       1 reflector.go:205] github.com/coredns/coredns/plugin/kubernetes/controller.go:313: Failed to list *v1.Endpoints: Get https://10.96.0.1:443/api/v1/endpoints?limit=500&resourceVersion=0: dial tcp 10.96.0.1:443: i/o timeout

我尝试启动一个简单的alpinepod/容器。并且确实10.96.0.1不响应 ping 或其他任何东西。

我被困在这里。我尝试了很多谷歌,但没有任何结果,我的理解非常基本。我猜网络出了点问题,但我不知道是什么(对我来说,这样做似乎很可疑,get podspod 会显示主机 IP,但也许这也是正常的?)

标签: kubernetesinstallation

解决方案


我发现问题是由主机的iptables规则引起的。


推荐阅读