说明
本文系搭建kubernetes v1.18.5 集群笔记，使用三台虚拟机作为 CentOS 测试机，安装kubeadm、kubelet、kubectl均使用yum安装，网络组件选用的是 flannel
行文中难免出现错误，如果读者有高见，请评论与我交流
如需转载请注明原始出处 https://www.cnblogs.com/hellxz/p/use-kubeadm-init-kubernetes-cluster.html
环境准备
部署集群没有特殊说明均使用root用户执行命令
硬件信息
| ip | hostname | mem | disk | explain |
|---|---|---|---|---|
| 192.168.87.145 | kube-master | 4 GB | 20GB | k8s 控制平台节点 |
| 192.168.87.146 | kube-node1 | 4 GB | 20GB | k8s 执行节点1 |
| 192.168.87.147 | kube-node2 | 4 GB | 20GB | k8s 执行节点2 |
软件信息
| software | version |
|---|---|
| CentOS | CentOS Linux release 7.7.1908 (Core) |
| Kubernetes | v1.18.5 |
| Docker | 19.03.12 |
保证环境正确性
| purpose | commands |
|---|---|
| 保证集群各节点互通 | ping -c 3 <ip> |
| 保证MAC地址唯一 | ip link 或 ifconfig -a |
| 保证集群内主机名唯一 | 查询 hostnamectl status，修改 hostnamectl set-hostname <hostname> |
| 保证系统产品uuid唯一 | dmidecode -s system-uuid 或 sudo cat /sys/class/dmi/id/product_uuid |
修改MAC地址参考命令：
ifconfig eth0 down ifconfig eth0 hw ether 00:0C:18:EF:FF:ED ifconfig eth0 up如product_uuid不唯一，请考虑重装CentOS系统
确保端口开放正常
kube-master节点端口检查：
| Protocol | Direction | Port Range | Purpose |
|---|---|---|---|
| TCP | Inbound | 6443* | kube-api-server |
| TCP | Inbound | 2379-2380 | etcd API |
| TCP | Inbound | 10250 | Kubelet API |
| TCP | Inbound | 10251 | kube-scheduler |
| TCP | Inbound | 10252 | kube-controller-manager |
kube-node*节点端口检查：
| Protocol | Direction | Port Range | Purpose |
|---|---|---|---|
| TCP | Inbound | 10250 | Kubelet API |
| TCP | Inbound | 30000-32767 | NodePort Services |
如果你对主机的防火墙配置不是很自信，可以关掉防火墙：
systemctl disable --now firewalld或者 清除iptables规则 （慎用）
iptables -F
配置主机互信
分别在各节点配置hosts映射：
cat >> /etc/hosts <<EOF
192.168.87.145 kube-master
192.168.87.146 kube-node1
192.168.87.147 kube-node2
EOF
kube-master生成ssh密钥，分发公钥到各节点：
#生成ssh密钥，直接一路回车
ssh-keygen -t rsa
#复制刚刚生成的密钥到各节点可信列表中，需分别输入各主机密码
ssh-copy-id root@kube-master
ssh-copy-id root@kube-node1
ssh-copy-id root@kube-node2
禁用swap
swap仅当内存不够时会使用硬盘块充当额外内存，硬盘的io较内存差距极大，禁用swap以提高性能
各节点均需执行：
swapoff -a
sed -i 's/.*swap.*/#&/' /etc/fstab
关闭 SELinux
关闭 SELinux，否则 kubelet 挂载目录时可能报错 Permission denied，可以设置为permissive或disabled，permissive 会提示warn信息
各节点均需执行：
setenforce 0
sed -i 's/^SELINUX=.*/SELINUX=disabled/' /etc/selinux/config
设置系统时区、同步时间
timedatectl set-timezone Asia/Shanghai
systemctl enable --now chronyd
查看同步状态：
timedatectl status
输出：
System clock synchronized: yes
NTP service: active
RTC in local TZ: no
System clock synchronized: yes，表示时钟已同步；NTP service: active，表示开启了时钟同步服务；
# 将当前的 UTC 时间写入硬件时钟
timedatectl set-local-rtc 0
# 重启依赖于系统时间的服务
systemctl restart rsyslog && systemctl restart crond
部署docker
所有节点均需安装部署docker
添加docker yum源
#安装必要依赖
yum install -y yum-utils device-mapper-persistent-data lvm2
#添加aliyun docker-ce yum源
yum-config-manager --add-repo http://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo
#重建yum缓存
yum makecache fast
安装docker
#查看可用docker版本
yum list docker-ce.x86_64 --showduplicates | sort -r
#安装指定版本docker
yum install -y docker-ce-19.03.12-3.el7
这里以安装19.03.12版本举例，注意版本号不包含
:与之前的数字
确保网络模块开机自动加载
lsmod | grep overlay
lsmod | grep br_netfilter
若上面命令无返回值输出或提示文件不存在，需执行以下命令：
cat > /etc/modules-load.d/docker.conf <<EOF
overlay
br_netfilter
EOF
modprobe overlay
modprobe br_netfilter
使桥接流量对iptables可见
各节点均需执行：
cat > /etc/sysctl.d/k8s.conf <<EOF
net.bridge.bridge-nf-call-ip6tables = 1
net.bridge.bridge-nf-call-iptables = 1
EOF
sysctl --system
验证是否生效，均返回 1 即正确
sysctl -n net.bridge.bridge-nf-call-iptables
sysctl -n net.bridge.bridge-nf-call-ip6tables
配置docker
mkdir /etc/docker
#修改cgroup驱动为systemd[k8s官方推荐]、限制容器日志量、修改存储类型，最后的docker家目录可修改
cat > /etc/docker/daemon.json <<EOF
{
"exec-opts": ["native.cgroupdriver=systemd"],
"log-driver": "json-file",
"log-opts": {
"max-size": "100m"
},
"storage-driver": "overlay2",
"storage-opts": [
"overlay2.override_kernel_check=true"
],
"registry-mirrors": ["https://7uuu3esz.mirror.aliyuncs.com"],
"data-root": "/data/docker"
}
EOF
#添加开机自启，立即启动
systemctl enable --now docker
验证docker是否正常
#查看docker信息，判断是否与配置一致
docker info
#hello-docker测试
docker run --rm hello-world
#删除测试image
docker rmi hello-world
添加用户到docker组
非root用户，无需sudo即可使用docker命令
#添加用户到docker组
usermod -aG docker <USERNAME>
#当前会话立即更新docker组
newgrp docker
部署kubernetes集群
未特殊说明，各节点均需执行如下步骤
添加kubernetes源
cat > /etc/yum.repos.d/kubernetes.repo <<EOF
[kubernetes]
name=Kubernetes
baseurl=https://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64/
enabled=1
gpgcheck=1
repo_gpgcheck=1
gpgkey=https://mirrors.aliyun.com/kubernetes/yum/doc/yum-key.gpg https://mirrors.aliyun.com/kubernetes/yum/doc/rpm-package-key.gpg
EOF
#重建yum缓存，输入y添加证书认证
yum makecache fast
安装kubeadm、kubelet、kubectl
各节点均需安装kubeadm、kubelet，kubectl仅kube-master节点需安装（作为worker节点，kubectl无法使用，可以不装）
yum install -y kubelet kubeadm kubectl --disableexcludes=kubernetes
systemctl enable --now kubelet
配置自动补全命令
#安装bash自动补全插件
yum install bash-completion -y
#设置kubectl与kubeadm命令补全，下次login生效
kubectl completion bash >/etc/bash_completion.d/kubectl
kubeadm completion bash > /etc/bash_completion.d/kubeadm
预拉取kubernetes镜像
由于国内网络因素，kubernetes镜像需要从mirrors站点或通过dockerhub用户推送的镜像拉取
#查看指定k8s版本需要哪些镜像
kubeadm config images list --kubernetes-version v1.18.5
另因阿里云的镜像暂时还没更新到v1.18.5版本，所以通过dockerhub上拉取，目前阿里云最新同步版本是v1.18.3，想通过v1.18.3版本拉取镜像请参考 <https://www.cnblogs.com/hellxz/p/13204093.html
在 /root/k8s 目录下，新建脚本get-k8s-images.sh，内容如下：
#!/bin/bash
# Script For Quick Pull K8S Docker Images
# by Hellxz Zhang <hellxz001@foxmail.com>
KUBE_VERSION=v1.18.5
PAUSE_VERSION=3.2
CORE_DNS_VERSION=1.6.7
ETCD_VERSION=3.4.3-0
# pull kubernetes images from hub.docker.com
docker pull kubeimage/kube-proxy-amd64:$KUBE_VERSION
docker pull kubeimage/kube-controller-manager-amd64:$KUBE_VERSION
docker pull kubeimage/kube-apiserver-amd64:$KUBE_VERSION
docker pull kubeimage/kube-scheduler-amd64:$KUBE_VERSION
# pull aliyuncs mirror docker images
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/pause:$PAUSE_VERSION
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/coredns:$CORE_DNS_VERSION
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/etcd:$ETCD_VERSION
# retag to k8s.gcr.io prefix
docker tag kubeimage/kube-proxy-amd64:$KUBE_VERSION k8s.gcr.io/kube-proxy:$KUBE_VERSION
docker tag kubeimage/kube-controller-manager-amd64:$KUBE_VERSION k8s.gcr.io/kube-controller-manager:$KUBE_VERSION
docker tag kubeimage/kube-apiserver-amd64:$KUBE_VERSION k8s.gcr.io/kube-apiserver:$KUBE_VERSION
docker tag kubeimage/kube-scheduler-amd64:$KUBE_VERSION k8s.gcr.io/kube-scheduler:$KUBE_VERSION
docker tag registry.cn-hangzhou.aliyuncs.com/google_containers/pause:$PAUSE_VERSION k8s.gcr.io/pause:$PAUSE_VERSION
docker tag registry.cn-hangzhou.aliyuncs.com/google_containers/coredns:$CORE_DNS_VERSION k8s.gcr.io/coredns:$CORE_DNS_VERSION
docker tag registry.cn-hangzhou.aliyuncs.com/google_containers/etcd:$ETCD_VERSION k8s.gcr.io/etcd:$ETCD_VERSION
# untag origin tag, the images won't be delete.
docker rmi kubeimage/kube-proxy-amd64:$KUBE_VERSION
docker rmi kubeimage/kube-controller-manager-amd64:$KUBE_VERSION
docker rmi kubeimage/kube-apiserver-amd64:$KUBE_VERSION
docker rmi kubeimage/kube-scheduler-amd64:$KUBE_VERSION
docker rmi registry.cn-hangzhou.aliyuncs.com/google_containers/pause:$PAUSE_VERSION
docker rmi registry.cn-hangzhou.aliyuncs.com/google_containers/coredns:$CORE_DNS_VERSION
docker rmi registry.cn-hangzhou.aliyuncs.com/google_containers/etcd:$ETCD_VERSION
脚本添加可执行权限，执行脚本拉取镜像：
chmod +x get-k8s-images.sh
./get-k8s-images.sh
拉取完成，执行 docker images 查看镜像
初始化kube-master
仅 kube-master 节点需要执行此步骤
修改kubelet配置默认cgroup driver
cat > /var/lib/kubelet/config.yaml <<EOF
apiVersion: kubelet.config.k8s.io/v1beta1
kind: KubeletConfiguration
cgroupDriver: systemd
EOF
systemctl restart kubelet
生成kubeadm初始化配置文件 [可选] 仅当需自定义初始化配置时用
kubeadm config print init-defaults > init.default.yaml
测试环境是否正常（WARNING是正常的）
kubeadm init phase preflight [--config kubeadm-init.yaml]
上图提示Warning是正常的，校验不了k8s信息是因为连不上被ban的网站，最后一个提示是因我本地未关闭防火墙，请我看清楚必要放行的端口号是否畅通
初始化master 10.244.0.0/16是flannel固定使用的IP段，设置取决于网络组件要求
kubeadm init --pod-network-cidr=10.244.0.0/16 --kubernetes-version=v1.18.5 [--config kubeadm-init.yaml]
输出如下：
[root@kube-master k8s]# kubeadm init --pod-network-cidr=10.244.0.0/16 --kubernetes-version=v1.18.5
W0703 18:49:19.076654 16469 configset.go:202] WARNING: kubeadm cannot validate component configs for API groups [kubelet.config.k8s.io kubeproxy.config.k8s.io]
[init] Using Kubernetes version: v1.18.5
[preflight] Running pre-flight checks
[WARNING Firewalld]: firewalld is active, please ensure ports [6443 10250] are open or your cluster may not function correctly
[preflight] Pulling images required for setting up a Kubernetes cluster
[preflight] This might take a minute or two, depending on the speed of your internet connection
[preflight] You can also perform this action in beforehand using 'kubeadm config images pull'
[kubelet-start] Writing kubelet environment file with flags to file "/var/lib/kubelet/kubeadm-flags.env"
[kubelet-start] Writing kubelet configuration to file "/var/lib/kubelet/config.yaml"
[kubelet-start] Starting the kubelet
[certs] Using certificateDir folder "/etc/kubernetes/pki"
[certs] Generating "ca" certificate and key
[certs] Generating "apiserver" certificate and key
[certs] apiserver serving cert is signed for DNS names [kube-master kubernetes kubernetes.default kubernetes.default.svc kubernetes.default.svc.cluster.local] and IPs [10.96.0.1 192.168.87.145]
[certs] Generating "apiserver-kubelet-client" certificate and key
[certs] Generating "front-proxy-ca" certificate and key
[certs] Generating "front-proxy-client" certificate and key
[certs] Generating "etcd/ca" certificate and key
[certs] Generating "etcd/server" certificate and key
[certs] etcd/server serving cert is signed for DNS names [kube-master localhost] and IPs [192.168.87.145 127.0.0.1 ::1]
[certs] Generating "etcd/peer" certificate and key
[certs] etcd/peer serving cert is signed for DNS names [kube-master localhost] and IPs [192.168.87.145 127.0.0.1 ::1]
[certs] Generating "etcd/healthcheck-client" certificate and key
[certs] Generating "apiserver-etcd-client" certificate and key
[certs] Generating "sa" key and public key
[kubeconfig] Using kubeconfig folder "/etc/kubernetes"
[kubeconfig] Writing "admin.conf" kubeconfig file
[kubeconfig] Writing "kubelet.conf" kubeconfig file
[kubeconfig] Writing "controller-manager.conf" kubeconfig file
[kubeconfig] Writing "scheduler.conf" kubeconfig file
[control-plane] Using manifest folder "/etc/kubernetes/manifests"
[control-plane] Creating static Pod manifest for "kube-apiserver"
[control-plane] Creating static Pod manifest for "kube-controller-manager"
W0703 18:49:23.039913 16469 manifests.go:225] the default kube-apiserver authorization-mode is "Node,RBAC"; using "Node,RBAC"
[control-plane] Creating static Pod manifest for "kube-scheduler"
W0703 18:49:23.040907 16469 manifests.go:225] the default kube-apiserver authorization-mode is "Node,RBAC"; using "Node,RBAC"
[etcd] Creating static Pod manifest for local etcd in "/etc/kubernetes/manifests"
[wait-control-plane] Waiting for the kubelet to boot up the control plane as static Pods from directory "/etc/kubernetes/manifests". This can take up to 4m0s
[apiclient] All control plane components are healthy after 21.505101 seconds
[upload-config] Storing the configuration used in ConfigMap "kubeadm-config" in the "kube-system" Namespace
[kubelet] Creating a ConfigMap "kubelet-config-1.18" in namespace kube-system with the configuration for the kubelets in the cluster
[upload-certs] Skipping phase. Please see --upload-certs
[mark-control-plane] Marking the node kube-master as control-plane by adding the label "node-role.kubernetes.io/master=''"
[mark-control-plane] Marking the node kube-master as control-plane by adding the taints [node-role.kubernetes.io/master:NoSchedule]
[bootstrap-token] Using token: 2b7cfv.6bhz4z3a3vzyg498
[bootstrap-token] Configuring bootstrap tokens, cluster-info ConfigMap, RBAC Roles
[bootstrap-token] configured RBAC rules to allow Node Bootstrap tokens to get nodes
[bootstrap-token] configured RBAC rules to allow Node Bootstrap tokens to post CSRs in order for nodes to get long term certificate credentials
[bootstrap-token] configured RBAC rules to allow the csrapprover controller automatically approve CSRs from a Node Bootstrap Token
[bootstrap-token] configured RBAC rules to allow certificate rotation for all node client certificates in the cluster
[bootstrap-token] Creating the "cluster-info" ConfigMap in the "kube-public" namespace
[kubelet-finalize] Updating "/etc/kubernetes/kubelet.conf" to point to a rotatable kubelet client certificate and key
[addons] Applied essential addon: CoreDNS
[addons] Applied essential addon: kube-proxy
Your Kubernetes control-plane has initialized successfully!
To start using your cluster, you need to run the following as a regular user:
mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config
You should now deploy a pod network to the cluster.
Run "kubectl apply -f [podnetwork].yaml" with one of the options listed at:
https://kubernetes.io/docs/concepts/cluster-administration/addons/
Then you can join any number of worker nodes by running the following on each as root:
kubeadm join 192.168.87.145:6443 --token 2b7cfv.6bhz4z3a3vzyg498
--discovery-token-ca-cert-hash sha256:79bd63d82634f9953cc9d6b5a923fa87c973f0c3fd9ed7270167052dd834c026
为日常使用集群的用户添加kubectl使用权限
su hellxz
mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/admin.conf
sudo chown $(id -u):$(id -g) $HOME/.kube/admin.conf
echo "export KUBECONFIG=$HOME/.kube/admin.conf" >> ~/.bashrc
exit
配置master认证
echo 'export KUBECONFIG=/etc/kubernetes/admin.conf' >> /etc/profile
. /etc/profile
如果不配置这个，会提示如下输出：
The connection to the server localhost:8080 was refused - did you specify the right host or port?此时master节点已经初始化成功，但是还未完装网络组件，还无法与其他节点通讯
安装网络组件，以flannel为例
cd ~/k8s
yum install -y wget
#下载flannel最新配置文件
wget https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml
kubectl apply -f kube-flannel.yml
查看kube-master节点状态
kubectl get nodes
如果STATUS提示
NotReady，可以通过kubectl describe node kube-master查看具体的描述信息，性能差的服务器到达Ready状态时间会长些
备份镜像供其他节点使用
在kube-master节点将镜像备份出来，便于后续传输给其他node节点，当然有镜像仓库更好
docker save k8s.gcr.io/kube-proxy:v1.18.5
k8s.gcr.io/kube-apiserver:v1.18.5
k8s.gcr.io/kube-controller-manager:v1.18.5
k8s.gcr.io/kube-scheduler:v1.18.5
k8s.gcr.io/pause:3.2
k8s.gcr.io/coredns:1.6.7
k8s.gcr.io/etcd:3.4.3-0 > k8s-imagesV1.18.5.tar
初始化kube-node*节点并加入集群
拷贝镜像到node节点，以kube-node1举例，node2不再累述
#此时命令在kube-node*节点上执行
mkdir ~/k8s
scp root@kube-master:/root/k8s/k8s-imagesV1.18.5.tar ~/k8s
获取加入kubernetes命令，未忘可不选
刚才在初始化kube-master节点时，有在最后输出其加入集群的命令，假如我没记下来，那怎么办呢？
访问kube-master输入创建新token命令，同时输出加入集群的命令：
kubeadm token create --print-join-command
在kube-node*节点上执行加入集群命令
kubeadm join 192.168.87.145:6443 --token jdyzyq.icwlpkm36kgs6nqh --discovery-token-ca-cert-hash sha256:24f9b05fa10307ef6fff4132e0ec3c8b54917d4ff440b36108908aca588d8be7
查看集群节点状态
kubectl get nodes
参考
- 《Kubernetes权威指南》第4版
- 官方文档 https://kubernetes.io/docs/setup/production-environment/tools/kubeadm/install-kubeadm/
至此，本文结束，感谢阅读，如果对你有帮助，欢迎点推荐，如果有问题，请在下方留言。