多集群的切换是K8s运维中比不可少的问题,常见的基于多个集群进行切换的方法有三种:
- 切换config文件
- 通过context进行集群切换
- 用kubecm进行集群切换
切换config文件
我们先看看放在.kube下默认的config集群:
$ kubectl config view apiVersion: v1 clusters: - cluster: certificate-authority-data: DATA+OMITTED server: name: kubernetes contexts: - context: cluster: kubernetes user: "2xxxxxxxxx7" name: 2xxxxxxxxxxxxxxxx8 current-context: 2xxxxxxxxxxxx8 kind: Config preferences: {} users: - name: "2xxxxxxxxxxx27" user: client-certificate-data: client-key-data:
通过export更改默认集群配置:
$ export KUBECONFIG=$HOME/.kube/rancher-config
查看:
$ kubectl config view apiVersion: v1 clusters: - cluster: certificate-authority-data: DATA+OMITTED server: name: stage contexts: - context: cluster: stage user: stage name: stage current-context: stage kind: Config preferences: {} users: - name: stage user: token:
这种方法就是特别麻烦,每次切换都要通过export设置,非常麻烦,同时也不方便。
通过context进行集群切换
把要切换的所有config文件添加到KUBECONFIG环境变量中
$ export KUBECONFIG=$HOME/.kube/config:$HOME/.kube/rancher-config
或者可以将命令直接写到.bashrc文件中,这样就不需要每次手动设置。
通过kubectl config view可以看到这两个文件已经合并到一起了:
$ kubectl config view apiVersion: v1 clusters: - cluster: certificate-authority-data: DATA+OMITTED server: name: kubernetes contexts: - context: cluster: kubernetes user: "2xxxxxxxxx7" name: 2xxxxxxxxxxxxxxxx8 current-context: 2xxxxxxxxxxxx8 kind: Config preferences: {} users: - name: "2xxxxxxxxxxx27" user: client-certificate-data: client-key-data:
合并一起后就可以通过use-context切换上下文来切换集群:
# 切换stage集群 $ kubectl config use-context stage Switched to context "stage".
切换到第二个集群:
$ kubectl config use-context 2xxxxxxxxxxx8 Switched to context "21xxxxxxxxxxxxx9648".
但这种方法不方便的地方就在如果context很多的时候,不方便查看。
用kubecm进行集群切换
安装篇
基础环境
mac
k8s 1.16.9
# wget https://github.com.cnpmjs.org/sunny0826/kubecm/releases/download/v0.10.3/kubecm_0.10.3_Darwin_x86_64.tar.gz # tar xf kubecm_0.10.3_Darwin_x86_64.tar.gz # mv kubecm /usr/local/bin # kubecm --help KubeConfig Manager _ _ | | ___ _| |__ ___ ___ _ __ ___ | |/ / | | | '_ \ / _ \/ __| '_ \ _ \ | <| |_| | |_) | __/ (__| | | | | | |_|\_\\__,_|_.__/ \___|\___|_| |_| |_| Find more information at: https://github.com/sunny0826/kubecm Usage: kubecm [command] Available Commands: add Merge configuration file with $HOME/.kube/config alias Generate alias for all contexts completion Generates bash/zsh completion scripts delete Delete the specified context from the kubeconfig help Help about any command ls List kubeconfig merge Merge the kubeconfig files in the specified directory namespace Switch or change namespace interactively rename Rename the contexts of kubeconfig switch Switch Kube Context interactively version Print version info Flags: --config string path of kubeconfig (default "/var/root/.kube/config") -h, --help help for kubecm Use "kubecm [command] --help" for more information about a command.
补全
bash
# ource > ~/.bashrc # source ~/.bashrc
zsh
# source > ~/.zshrc # source ~/.zshrc
基本使用
查看
# kubecm ls +------------+-----------------------------------+---------------+-----------------------+--------------------------------+--------------+ | CURRENT | NAME | CLUSTER | USER | SERVER | Namespace | +============+===================================+===============+=======================+================================+==============+ | * | 283387751775489594-c02e82b6ada | kubernetes | 283387751775489594 | https://3.3.3.3:6443 | default | | | f044bd94d251fea1279648 | | | | | +------------+-----------------------------------+---------------+-----------------------+--------------------------------+--------------+ Cluster check succeeded! Kubernetes master is running at https://3.3.3.3:6443 Contains components: [scheduler controller-manager etcd-5 etcd-0 etcd-4 etcd-1 etcd-3 etcd-2]
添加
# ll -rw-r--r-- 1 root wheel 5.9K 8 26 15:39 config.shengchan -rw-r--r-- 1 root wheel 5.9K 11 18 15:59 config.yufa # kubecm add -f config.yufa -c -n stage # kubecm add -f config.shengchan -c -n production # kubecm ls +------------+---------------+-----------------------+--------------------+--------------------------------+--------------+ | CURRENT | NAME | CLUSTER | USER | SERVER | Namespace | +============+===============+=======================+====================+================================+==============+ | * | production | cluster-kg7h48f8tb | user-kg7h48f8tb | https://3.3.3.3:6443 | default | +------------+---------------+-----------------------+--------------------+--------------------------------+--------------+ | | stage | cluster-kfc6fbfcdt | user-kfc6fbfcdt | https://4.4.4.4:6443 | default | +------------+---------------+-----------------------+--------------------+--------------------------------+--------------+
删除
# kubecm delete ???? Select:production ???? True Context Delete:「production」 「/var/root/.kube/config」 write successful! +------------+---------------+-----------------------+--------------------+--------------------------------+--------------+ | CURRENT | NAME | CLUSTER | USER | SERVER | Namespace | +============+===============+=======================+====================+================================+==============+ | * | stage | cluster-kfc6fbfcdt | user-kfc6fbfcdt | https://4.4.4.4:6443 | default | +------------+---------------+-----------------------+--------------------+--------------------------------+--------------+
切换
# kubecm switch ???? Select:production 「/var/root/.kube/config」 write successful! +------------+---------------+-----------------------+--------------------+--------------------------------+--------------+ | CURRENT | NAME | CLUSTER | USER | SERVER | Namespace | +============+===============+=======================+====================+================================+==============+ | * | production | cluster-kg7h48f8tb | user-kg7h48f8tb | https://3.3.3.3:6443 | default | +------------+---------------+-----------------------+--------------------+--------------------------------+--------------+ | | stage | cluster-kfc6fbfcdt | user-kfc6fbfcdt | https://4.4.4.4:6443 | default | +------------+---------------+-----------------------+--------------------+--------------------------------+--------------+ Switched to context 「production」