资源

节点名称	IP	配置	系统
node3	10.2.20.17	4核8G	Ubuntu Server 22.04 LTS 64bit
node2	10.2.24.4	4核8G	Ubuntu Server 22.04 LTS 64bit
node1	10.2.20.13	4核8G	Ubuntu Server 22.04 LTS 64bit
master	10.2.24.10	4核8G	Ubuntu Server 22.04 LTS 64bit

ps:所有命令尽量使用root账号操作

1.设置节点名称，留一台当做master

#newhostname改为节点名称，每台服务器都要执行，规划好名称，名称会在k8s显示。
sudo hostnamectl set-hostname [newhostname]

2.关闭swap，selinux，firewalld

# 临时关闭swap
sudo swapoff -a
# 永久关闭swap,打开后找到其中未注释的代码，如下图所示
sudo vi /etc/fstab

https://i-blog.csdnimg.cn/direct/ea56babcbaf645388794e995941238f8.png

# 临时关闭selinux
sudo setenforce 0
# 永久关闭selinux
sudo vi /etc/selinux/config
# 将SELINUX=enforcing改为SELINUX=disabled，保存
# 关闭防火墙
sudo systemctl stop ufw
sudo systemctl disable ufw

3.在每台服务器上执行命令安装microk8s

# 最新版本安装,选择一种，不要两个都执行
sudo snap install microk8s --classic
# 指定版本安装
sudo snap install microk8s --classic --channel=1.24/stable

安装成功后会有类似于如下输出

microk8s (1.25/stable) v1.25.2 from Canonical✓ installed

4.设置K8s 命令别名（alias）（仅在master上执行）

把一些常用的命令设置为别名，更方便
# 设置 microk8s.kubectl 命令为 kubectl
sudo snap alias microk8s.kubectl kubectl
# 设置成功后出现以下的输出，代表设置成功
Added:- microk8s.kubectl as kubectl
# 设置 microk8s.ctr 命令为 ctr
sudo snap alias microk8s.ctr ctr
ps:也可以设置其他的常用命令，设置方式类似

5.检查microk8s status状态及解决

ps:如果已启动则跳过该步骤

#执行状态查询
sudo microk8s status
#如果出现以下则表示未启动
sudo microk8s is not running. Use microk8s inspect for a deeper inspection.

我们可以执行microk8s.inspect查看是否有报错信息，如果没有任何报错，可以来通过kubectl来进一步排查问题到底出现在了哪里

sudo kubectl get pods --all-namespaces# 命令执行完毕，将得到类似下面的日志结果：
NAMESPACE     NAME                                       READY   STATUS     RESTARTS   AGE
kube-system   calico-node-nnshm                          0/1     Init:0/2   0          9m
kube-system   calico-kube-controllers-67774c44db-mcmvx   0/1     Pending    0          9m

使用kubectl describe pod来查看 STATUS 状态为 Init 的这个pod：

sudo kubectl describe pod calico-node-nnshm -n kube-system# 得到如下的输出：
Failed to create pod sandbox: rpc error: code = DeadlineExceeded desc = failed to get sandbox image 
"registry.k8s.io/pause:3.7": failed to pull image "registry.k8s.io/pause:3.7": failed to pull and unpack 
image "registry.k8s.io/pause:3.7": failed to resolve reference "registry.k8s.io/pause:3.7": failed to do 
request: Head "https://asia-east1-docker.pkg.dev/v2/k8s-artifacts-prod/images/pause/manifests/3.7": dial 
tcp 142.250.157.82:443: i/o timeout

ps:如果报错是上面这个,则切换未阿里云数据源

sudo ctr image pull registry.cn-hangzhou.aliyuncs.com/google_containers/pause:3.7# 拉取成功： 提示如下
registry.cn-hangzhou.aliyuncs.com/google_containers/pause:3.7:                    resolved       |++++++++++++++++++++++++++++++++++++++| 
index-sha256:bb6ed397957e9ca7c65ada0db5c5d1c707c9c8afc80a94acbe69f3ae76988f0c:    done           |++++++++++++++++++++++++++++++++++++++| 
manifest-sha256:f81611a21cf91214c1ea751c5b525931a0e2ebabe62b3937b6158039ff6f922d: done           |++++++++++++++++++++++++++++++++++++++| 
layer-sha256:7582c2cc65ef30105b84c1c6812f71c8012663c6352b01fe2f483238313ab0ed:    done           |++++++++++++++++++++++++++++++++++++++| 
config-sha256:221177c6082a88ea4f6240ab2450d540955ac6f4d5454f0e15751b653ebda165:   done           |++++++++++++++++++++++++++++++++++++++| 
elapsed: 1.8 s                                                                    total:  304.0  (168.8 KiB/s)                                     
unpacking linux/amd64 sha256:bb6ed397957e9ca7c65ada0db5c5d1c707c9c8afc80a94acbe69f3ae76988f0c...
done: 26.431318ms

然后更改tag，把registry.cn-hangzhou.aliyuncs.com/google_containers/pause:3.7镜像更改为registry.k8s.io/pause:3.7

sudo ctr image tag registry.cn-hangzhou.aliyuncs.com/google_containers/pause:3.7 registry.k8s.io/pause:3.7# 重启 microk8s
sudo microk8s stop && microk8s start# 重启后在查看 microk8s 的状态
sudo microk8s statusmicrok8s is running
high-availability: nodatastore master nodes: 127.0.0.1:19001datastore standby nodes: none
addons:enabled:ha-cluster           # (core) Configure high availability on the current nodehelm                 # (core) Helm - the package manager for Kuberneteshelm3                # (core) Helm 3 - the package manager for Kubernetesdisabled:cert-manager         # (core) Cloud native certificate managementcommunity            # (core) The community addons repositorydashboard            # (core) The Kubernetes dashboarddns                  # (core) CoreDNSgpu                  # (core) Automatic enablement of Nvidia CUDAhost-access          # (core) Allow Pods connecting to Host services smoothlyhostpath-storage     # (core) Storage class; allocates storage from host directoryingress              # (core) Ingress controller for external accesskube-ovn             # (core) An advanced network fabric for Kubernetesmayastor             # (core) OpenEBS MayaStormetallb              # (core) Loadbalancer for your Kubernetes clustermetrics-server       # (core) K8s Metrics Server for API access to service metricsminio                # (core) MinIO object storageobservability        # (core) A lightweight observability stack for logs, traces and metricsprometheus           # (core) Prometheus operator for monitoring and loggingrbac                 # (core) Role-Based Access Control for authorisationregistry             # (core) Private image registry exposed on localhost:32000storage              # (core) Alias to hostpath-storage add-on, deprecated

如果还是没启起来，再次查看pod的状态，如果显示以下的错误，代表无法拉取calico插件，需要去github上面手动拉取 https://github.com/projectcalico/calico

下载对应的版本包，下载完成后，把images目录上传到服务器上，导入需要的镜像包即可，我这里下载的3.25.1
也可以从百度网盘下载，链接: https://pan.baidu.com/s/1tEodKVwFEtfQwYz2UW3MQg 提取码: 2336

上传至任意目录，然后进入该目录执行以下命令

sudo ctr images import calico-kube-controllers.tar && ctr images import calico-cni.tar && ctr images import calico-node.tar

成功后显示

6.拉取coredns的国内镜像

sudo ctr images pull registry.aliyuncs.com/google_containers/coredns:1.10.1
# 修改coredns的镜像
sudo kubectl edit deploy coredns -n kube-system
# 把镜像修改成 registry.aliyuncs.com/google_containers/coredns:1.10.1

7.配置 K8s 集群
7.1 先修改 master 服务器的hosts文件

sudo vi /etc/hosts# 把 master 的节点IP和hostname以及所有 node 节点的IP和hostname写进去
127.0.1.1 localhost.localdomain VM-0-8-ubuntu
127.0.0.1 localhost
10.2.24.10 master       # master 的节点IP和hostname
110.2.20.13 node1        # node 节点的IP和hostname
10.2.24.4 node2
10.2.20.17 node3::1 ip6-localhost ip6-loopback
fe00::0 ip6-localnet
ff00::0 ip6-mcastprefix
ff02::1 ip6-allnodes
ff02::2 ip6-allrouters
ff02::3 ip6-allhosts

7.2 所有 node 服务器的hosts文件都需要更改

sudo vi /etc/hosts# 把当前 node 节点的IP和hostname写进去
127.0.1.1 localhost.localdomain VM-0-6-ubuntu
127.0.0.1 localhost
110.2.20.13 node1::1 ip6-localhost ip6-loopback
fe00::0 ip6-localnet
ff00::0 ip6-mcastprefix
ff02::1 ip6-allnodes
ff02::2 ip6-allrouters
ff02::3 ip6-allhosts

7.3 在master服务器上执行

sudo microk8s add-node
生成如下的命令
From the node you wish to join to this cluster, run the following:
microk8s join 10.2.24.10:25000/89f1088fa37a9e79da47556830a4ce77/f619bfa7e77cUse the '--worker' flag to join a node as a worker not running the control plane, eg:
microk8s join 10.2.24.10:25000/89f1088fa37a9e79da47556830a4ce77/f619bfa7e77c --worker   
# 复制这一条命令在 node 服务器上执行，需要注意的是生成的命令是一次性的，所以如果我们需要添加多个工作节点，需要重复在master中执行 microk8s add-node，获得带有不同的 token 参数的命令If the node you are adding is not reachable through the default interface you can use one of the following:
microk8s join 10.2.24.10:25000/89f1088fa37a9e79da47556830a4ce77/f619bfa7e77c

7.4 在node服务器上执行

#这条命令从7.3执行结果复制，每台node只能执行一次，且注意25000端口需要在云服务器开放，避免访问不通
#命令需要等待20秒左右后会成功
sudo microk8s join 10.2.24.10:25000/ebbe369047950c6e4262e82e14d4d0ef/f619bfa7e77c --worker
Contacting cluster at 172.27.0.8The node has joined the cluster and will appear in the nodes list in a few seconds.This worker node gets automatically configured with the API server endpoints.
If the API servers are behind a loadbalancer please set the '--refresh-interval' to '0s' in:/var/snap/microk8s/current/args/apiserver-proxy
and replace the API server endpoints with the one provided by the loadbalancer in:/var/snap/microk8s/current/args/traefik/provider.yaml

7.5 在master上检查所有节点的运行状况

sudo kubectl get pod -o=custom-columns=NAME:.metadata.name,STATUS:.status.phase,NODE:.spec.nodeName --all-namespaces
NAME                                       STATUS    NODE
calico-kube-controllers-67774c44db-mcmvx   Running   potato
calico-node-sgj2r                          Running   potato
calico-node-9pvcz                          Running   spud
calico-node-45kcb                          Running   murphy

7.6 开启microk8s的组件

# 开启dns组件
microk8s enable dns
# 开启rbac组件
microk8s enable rbac

7.7 安装kuboard-v3 ， 需要开启30080端口与10081端口

# 安装docker，安装依赖
sudo apt install apt-transport-https ca-certificates curl gnupg lsb-release
# 添加阿里云Docker镜像源GPG秘钥
sudo curl -fsSL https://mirrors.aliyun.com/docker-ce/linux/ubuntu/gpg | sudo gpg --dearmor -o /usr/share/keyrings/docker-archive-keyring.gpg
# 添加阿里云镜像源
sudo echo "deb [arch=amd64 signed-by=/usr/share/keyrings/docker-archive-keyring.gpg] https://mirrors.aliyun.com/docker-ce/linux/ubuntu $(lsb_release -cs) stable" | sudo tee /etc/apt/sources.list.d/docker.list > /dev/null
# 更新apt
sudo apt update
# 安装docker
sudo apt install docker-ce
# 安装kuboard
sudo docker run -d \--restart=unless-stopped \--name=kuboard \-p 30080:80/tcp \-p 10081:10081/tcp \-e KUBOARD_ENDPOINT="http://10.0.0.42:30080" \-e KUBOARD_AGENT_SERVER_TCP_PORT="10081" \-v /root/kuboard-data:/data \swr.cn-east-2.myhuaweicloud.com/kuboard/kuboard:v3

7.8 配置NFS

# 在master服务器上安装NFS服务器，执行命令安装NFS服务器
sudo apt install nfs-kernel-server
# 打开nfs服务器配置文件/etc/exports，指定nfs服务器共享目录及其属性, /mnt  *(rw,sync,no_root_squash)
sudo vi /etc/exports
# 重启NFS服务器
sudo service nfs-kernel-server restart
# 在所有node服务器上安装NFS客户端
sudo apt install nfs-common

7.9 给master添加污点以避免业务容器调度到master节点上（已有则忽略）。 如果允许master调度业务容器，那么这里就跳过

# 添加污点
sudo kubectl taint nodes master node-role.kubernetes.io/master=:NoSchedule

8.安装Python2

# 更新apt
sudo apt-get update
# 安装python2
sudo apt-get install python2
# 设置python替代方案
update-alternatives --install /usr/bin/python python /usr/bin/python2 1
update-alternatives --install /usr/bin/python python /usr/bin/python3 2
# 查看是否设置成功
update-alternatives --list python
# 切换python版本，输入python2对应的数字即可
update-alternatives --config python
# 查看切换是否成功
python --version

至此k8s搭建成功， 访问地址，master节点的公网ip加上端口第七部执行的docker启动端口30080
http://49.233.207.77:30080/kuboard/cluster
账号密码默认
admin
Kuboard123

k8s常见问题
1.访问k8s节点端口特别慢，在自己的节点上运行就返回很快：
将网卡的校验关闭即可

ethtool -K vxlan.calico tx-checksum-ip-generic off

2.Microk8s提示deployments.apps is forbidden: User “system:serviceaccount:kube-system:default” cannot list resource “deployments” in API group “apps” in the namespace “kube-system”

kubectl auth can-i list deployments --namespace kube-system --as <当前登录的用户名>

如果提示no则执行如下命令进行授权：

kubectl create clusterrolebinding my-cluster-admin --clusterrole=cluster-admin --serviceaccount=kube-system:<当前登录的用户名>

3.安装kuboardV3的时候执行 watch kubectl get pods -n kuboard 如果没有etcd的pod，就要手动去添加一个节点的etcd

# 默认在master上添加一个etcd
kubectl label nodes master node-role.kubernetes.io/master=
kubectl label nodes master node.kubernetes.io/microk8s-controlplane=microk8s-controlplane# 如果需要多个etcd，可以在其他的节点上在添加标签
kubectl label nodes 【your-node-name】 k8s.kuboard.cn/role=etcd
ps:etcd数量最好为奇数

4.证书延期

• 4.1 配置每半年自动执行一次更新（时间随意）

# 查看证书过期时间
microk8s.refresh-certs -c
# 延长ca证书
microk8s.refresh-certs -e ca.crt
# 延长server证书
microk8s.refresh-certs -e server.crt
# 延长front证书
microk8s.refresh-certs -e front-proxy-client.crt

把上面的命令写入一个sh脚本，例如：cert.sh

# 给sh脚本赋予执行权限
chmod +x cert.sh
# 配置定时任务
crontab -e
0 0 1 */6 * /mnt/cert.sh >> /mnt/cert.log

• 4.2 禁用snap自动更新

# 禁用自动更新
snap refresh --hold
# 启用自动更新
snap refresh --unhold