kubeadm join success but node not joined

发表于 分类于
本文字数： 1.5k 阅读时长 ≈ 1 分钟

问题

在搭建 Kubernetes 集群时，遇到这样一个问题，就是在 node 节点上使用 kubeadm join 时能够成功的加入节点，但是在 master 节点上却无法查看集群中的 node 节点。如下：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
node1$ sudo kubeadm join --token 1bc310.cb323487a828849e 10.2.7.114:6443 --discovery-token-ca-cert-hash sha256:3d39f8fe34a043ccef4821014fd6d3e0f222614d37d59a6e4944c74f257c6d4d
[preflight] Running pre-flight checks.
	[WARNING FileExisting-crictl]: crictl not found in system path
[discovery] Trying to connect to API Server "10.2.7.114:6443"
[discovery] Created cluster-info discovery client, requesting info from "https://10.2.7.114:6443"
[discovery] Requesting info from "https://10.2.7.114:6443" again to validate TLS against the pinned public key
[discovery] Cluster info signature and contents are valid and TLS certificate validates against pinned roots, will use API Server "10.2.7.114:6443"
[discovery] Successfully established connection with API Server "10.2.7.114:6443"

This node has joined the cluster:
* Certificate signing request was sent to master and a response
  was received.
* The Kubelet was informed of the new secure connection details.

Run 'kubectl get nodes' on the master to see this node join the cluster.
1
2
3
master$ kubectl get nodes
NAME      STATUS    ROLES     AGE       VERSION
ubuntu    Ready     master    4h        v1.9.1

解决方法

出现这个问题的原因是 node 节点的主机名与 master 节点的相同，因此需要给 所有的 node 节点取与 master 节点不同的主机名。

修改 /etc/hostname 以及 /etc/hosts 文件中的主机名，再通过命令临时设置主机名：sudo hostname 主机名。

配置完之后，在 node 节点上执行 kubeadm reset， 再重新执行 kubeadm join 。

最后在 master 节点上查看节点：

1
2
3
4
5
$ kubectl get nodes
NAME      STATUS    ROLES     AGE       VERSION
node1     Ready     <none>    1h        v1.9.1
node2     Ready     <none>    1h        v1.9.1
ubuntu    Ready     master    6h        v1.9.1

可以看到有两个 node 节点。

参考：issue 61224