first commit
This commit is contained in:
huanqing.shao
489
docs/install/install-kubernetes.md
Normal file
489
docs/install/install-kubernetes.md
Normal file
@ -0,0 +1,489 @@
|
||||
# 安装 Kubernetes
|
||||
|
||||
# 介绍
|
||||
|
||||
kubernetes 安装有多种选择，本文档描述的集群安装具备如下特点：
|
||||
|
||||
* 使用 kubernetes 1.15.0
|
||||
* 三个 master 组成主节点集群，通过内网 loader balancer 实现负载均衡
|
||||
* 多个 worker 组成工作节点集群，通过外网 loader balancer 实现负载均衡
|
||||
* 使用 calico 作为网络插件
|
||||
* 使用 nginx-ingress 作为 kubernetes ingress controller
|
||||
|
||||
[领取阿里云最高2000元红包](https://promotion.aliyun.com/ntms/yunparter/invite.html?userCode=obezo3pg)
|
||||
|
||||
下图为 kubernetes 安装的拓扑图
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
* 标准机用于制作 centos 镜像，该镜像预先安装了 docker / kubelet / kubectl / kubeadm， 并预先下载了 kubernetes 所需的 docker image
|
||||
* 跳板机用于通过 ssh 端口访问标准机、master 节点、worker 节点的 shell，如果您所使用的环境中，可以直接访问各节点的 shell，也可以无需跳板机
|
||||
* NAT y.y.y.y将 apiserver 的6443端口映射到外网，如果您始终只在 master 节点上执行 kubectl 命令，则无需做此映射
|
||||
* Load Balancer z.z.z.z 是 kubernetes 的 Ingress 外网映射，通常是必须的
|
||||
|
||||
|
||||
|
||||
# 安装步骤
|
||||
|
||||
|
||||
## 制作标准机镜像
|
||||
|
||||
通过使用标准机镜像，可以
|
||||
|
||||
- **避免重复执行对测试机安装必要软件的过程**
|
||||
- **以一种相对标准化的过程管理测试机的维护**
|
||||
|
||||
标准机镜像中预装了如下内容：
|
||||
|
||||
- docker
|
||||
- gitlab-runner
|
||||
- kubernetes images
|
||||
|
||||
本文档描述的安装过程已基于 centos 7.6 验证
|
||||
|
||||
标准机镜像的制作过程描述如下：
|
||||
|
||||
### 安装docker
|
||||
|
||||
**卸载旧版本**
|
||||
|
||||
```bash
|
||||
sudo yum remove docker \
|
||||
docker-client \
|
||||
docker-client-latest \
|
||||
docker-common \
|
||||
docker-latest \
|
||||
docker-latest-logrotate \
|
||||
docker-logrotate \
|
||||
docker-selinux \
|
||||
docker-engine-selinux \
|
||||
docker-engine
|
||||
```
|
||||
|
||||
**下载依赖包及安装包**
|
||||
|
||||
```bash
|
||||
wget https://download.docker.com/linux/centos/7/x86_64/stable/Packages/containerd.io-1.2.6-3.3.el7.x86_64.rpm
|
||||
|
||||
wget https://download.docker.com/linux/centos/7/x86_64/stable/Packages/docker-ce-cli-18.09.7-3.el7.x86_64.rpm
|
||||
|
||||
wget https://download.docker.com/linux/centos/7/x86_64/stable/Packages/docker-ce-18.09.7-3.el7.x86_64.rpm
|
||||
```
|
||||
|
||||
**安装**
|
||||
|
||||
```bash
|
||||
sudo yum install -y containerd.io-1.2.6-3.3.el7.x86_64.rpm
|
||||
sudo yum install -y docker-ce-cli-18.09.7-3.el7.x86_64.rpm
|
||||
sudo yum install -y docker-ce-18.09.7-3.el7.x86_64.rpm
|
||||
sudo systemctl enable docker
|
||||
```
|
||||
|
||||
**启动 docker 服务**
|
||||
|
||||
```bash
|
||||
sudo systemctl start docker
|
||||
```
|
||||
|
||||
**检查 docker 版本**
|
||||
|
||||
```bash
|
||||
docker version
|
||||
```
|
||||
|
||||
|
||||
|
||||
**参考文档**
|
||||
|
||||
https://docs.docker.com/install/linux/docker-ce/centos/
|
||||
|
||||
https://docs.docker.com/install/linux/linux-postinstall/
|
||||
|
||||
|
||||
|
||||
### 安装 nfs-utils
|
||||
|
||||
**执行安装命令**
|
||||
|
||||
```bash
|
||||
sudo yum install nfs-utils
|
||||
```
|
||||
|
||||
必须先安装 nfs-utils 才能挂载 nfs 网络存储
|
||||
|
||||
|
||||
|
||||
### K8S基本配置
|
||||
|
||||
**配置K8S的yum源**
|
||||
|
||||
```bash
|
||||
cat <<EOF > /etc/yum.repos.d/kubernetes.repo
|
||||
[kubernetes]
|
||||
name=Kubernetes
|
||||
baseurl=http://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64
|
||||
enabled=1
|
||||
gpgcheck=0
|
||||
repo_gpgcheck=0
|
||||
gpgkey=http://mirrors.aliyun.com/kubernetes/yum/doc/yum-key.gpg
|
||||
http://mirrors.aliyun.com/kubernetes/yum/doc/rpm-package-key.gpg
|
||||
EOF
|
||||
```
|
||||
|
||||
**关闭swap、防火墙**
|
||||
|
||||
```bash
|
||||
swapoff -a
|
||||
```
|
||||
|
||||
**关闭SeLinux**
|
||||
|
||||
```bash
|
||||
setenforce 0
|
||||
```
|
||||
|
||||
**修改 /etc/sysctl.conf**
|
||||
|
||||
```vim /etc/sysctl.conf```
|
||||
|
||||
向其中添加
|
||||
|
||||
```
|
||||
net.ipv4.ip_forward = 1
|
||||
net.bridge.bridge-nf-call-ip6tables = 1
|
||||
net.bridge.bridge-nf-call-iptables = 1
|
||||
```
|
||||
|
||||
如下图所示
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
**安装kubelet、kubeadm、kubectl**
|
||||
|
||||
```bash
|
||||
yum install -y kubelet-1.15.0 kubeadm-1.15.0 kubectl-1.15.0
|
||||
```
|
||||
|
||||
|
||||
|
||||
**修改docker Cgroup Driver为systemd**
|
||||
|
||||
```vim /usr/lib/systemd/system/docker.service```
|
||||
|
||||
向其中他添加
|
||||
|
||||
```--exec-opt native.cgroupdriver=systemd```
|
||||
|
||||
如下图所示
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
重启 docker
|
||||
|
||||
```
|
||||
systemctl daemon-reload
|
||||
systemctl restart docker
|
||||
```
|
||||
|
||||
|
||||
|
||||
**启动kubelet**
|
||||
|
||||
```bash
|
||||
systemctl enable kubelet && systemctl start kubelet
|
||||
```
|
||||
|
||||
|
||||
|
||||
**加载 kubernetes 镜像**
|
||||
|
||||
由于k8s服务相关镜像在国外镜像源，国内无法访问
|
||||
|
||||
执行以下命令添加docker k8s国内镜像源
|
||||
|
||||
```bash
|
||||
curl -sSL https://get.daocloud.io/daotools/set_mirror.sh | sh -s http://f1361db2.m.daocloud.io
|
||||
|
||||
systemctl restart docker
|
||||
```
|
||||
|
||||
**拉取k8s相关镜像**
|
||||
|
||||
```bash
|
||||
docker pull mirrorgooglecontainers/kube-apiserver:v1.15.0
|
||||
docker pull mirrorgooglecontainers/kube-controller-manager:v1.15.0
|
||||
docker pull mirrorgooglecontainers/kube-scheduler:v1.15.0
|
||||
docker pull mirrorgooglecontainers/kube-proxy:v1.15.0
|
||||
docker pull mirrorgooglecontainers/pause:3.1
|
||||
docker pull mirrorgooglecontainers/etcd:3.3.10
|
||||
docker pull coredns/coredns:1.3.1
|
||||
```
|
||||
|
||||
**更改镜像名为k8s官网镜像**
|
||||
|
||||
```bash
|
||||
docker tag d235b23c3570 k8s.gcr.io/kube-proxy:v1.15.0
|
||||
docker tag 201c7a840312 k8s.gcr.io/kube-apiserver:v1.15.0
|
||||
docker tag 2d3813851e87 k8s.gcr.io/kube-scheduler:v1.15.0
|
||||
docker tag 8328bb49b652 k8s.gcr.io/kube-controller-manager:v1.15.0
|
||||
docker tag da86e6ba6ca1 k8s.gcr.io/pause:3.1
|
||||
docker tag eb516548c180 k8s.gcr.io/coredns:1.3.1
|
||||
docker tag 2c4adeb21b4f k8s.gcr.io/etcd:3.3.10
|
||||
```
|
||||
|
||||
|
||||
|
||||
**制作镜像**
|
||||
|
||||
请参考阿里云基于ECS [制作虚拟机镜像](https://help.aliyun.com/document_detail/35109.html?spm=5176.2020520101.0.0.75fc4df5mtdFmV) 的文档
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
## 初始化API Server
|
||||
|
||||
### 创建 ApiServer 的 ELB（私网）
|
||||
|
||||
监听端口：6443 / TCP
|
||||
|
||||
后端资源组：包含 apple-master-a-1, apple-master-b-1, apple-master-b-2
|
||||
|
||||
后端端口：6443
|
||||
|
||||
开启 按源地址保持会话
|
||||
|
||||
假设完成创建以后，ELB的 ip 地址为 x.x.x.x
|
||||
|
||||
### 初始化第一个master节点
|
||||
|
||||
在 apple-master-a-1机器上执行
|
||||
|
||||
**配置 apiserver.apple 的域名**
|
||||
|
||||
```bash
|
||||
sudo -i
|
||||
echo "x.x.x.x apiserver.apple" >> /etc/hosts
|
||||
```
|
||||
|
||||
> 请替换其中的 x.x.x.x 为您的负载均衡服务器的实际 ip 地址
|
||||
|
||||
|
||||
|
||||
**创建 /root/k8s/kubeadm-config.yaml**
|
||||
|
||||
```yaml
|
||||
apiVersion: kubeadm.k8s.io/v1beta1
|
||||
kind: ClusterConfiguration
|
||||
kubernetesVersion: v1.14.3
|
||||
controlPlaneEndpoint: "apiserver.apple:6443"
|
||||
|
||||
```
|
||||
|
||||
|
||||
|
||||
**初始化 apiserver**
|
||||
|
||||
```bash
|
||||
kubeadm init --config=/root/k8s/kubeadm-config.yaml --upload-certs
|
||||
|
||||
```
|
||||
|
||||
执行结果如下图所示：
|
||||
|
||||
|
||||
|
||||
**初始化 gitlab-runner 用户的 kubectl 配置**
|
||||
|
||||
```bash
|
||||
rm -rf /home/gitlab-runner/.kube/
|
||||
mkdir /home/gitlab-runner/.kube/
|
||||
cp -i /etc/kubernetes/admin.conf /home/gitlab-runner/.kube/config
|
||||
chown -R gitlab-runner:gitlab-runner /home/gitlab-runner/.kube
|
||||
```
|
||||
|
||||
|
||||
|
||||
**安装 calico**
|
||||
|
||||
（需要在安全组–ServerFarm需要为集群服务器器端口6443建立安全组规则）
|
||||
|
||||
```bash
|
||||
su - gitlab-runner
|
||||
kubectl apply -f \
|
||||
https://docs.projectcalico.org/v3.6/getting-started/kubernetes/installation/hosted/kubernetes-datastore/calico-networking/1.7/calico.yaml
|
||||
```
|
||||
|
||||
> 安装calico， 请参考https://docs.projectcalico.org/v3.6/getting-started/kubernetes/
|
||||
|
||||
|
||||
|
||||
**等待calico安装就绪：**
|
||||
|
||||
执行如下命令，等待3-10分钟，直到所有的容器组处于 Running 状态
|
||||
|
||||
```sh
|
||||
watch kubectl get pod -n kube-system
|
||||
```
|
||||
|
||||
|
||||
|
||||
### 初始化第二、三个master节点
|
||||
|
||||
在 apple-master-b-1 和 apple-master-b-2 机器上执行
|
||||
|
||||
```bash
|
||||
sudo -i
|
||||
echo "x.x.x.x apiserver.apple" >> /etc/hosts
|
||||
```
|
||||
|
||||
|
||||
|
||||
执行 （以下命令行在初始化 apple-master-a-1 时，被打印在控制台上，执行时应该使用控制台打印出的命令参数）
|
||||
|
||||
```bash
|
||||
kubeadm join apiserver.apple:6443 --token ejwx62.vqwog6il5p83uk7y \
|
||||
--discovery-token-ca-cert-hash sha256:6f7a8e40a810323672de5eee6f4d19aa2dbdb38411845a1bf5dd63485c43d303 \
|
||||
--experimental-control-plane --certificate-key 70eb87e62f052d2d5de759969d5b42f372d0ad798f98df38f7fe73efdf63a13c
|
||||
```
|
||||
|
||||
|
||||
|
||||
### 检查 apiserver初始化结果
|
||||
|
||||
在第一个master节点 apple-master-a-1 上执行
|
||||
|
||||
```bash
|
||||
sudo -i
|
||||
su - gitlab-runner
|
||||
kubectl get nodes
|
||||
```
|
||||
|
||||
|
||||
|
||||
## 初始化 worker节点
|
||||
|
||||
### 获得 join命令参数
|
||||
|
||||
**在第一个master节点 apple-master-a-1 节点执行**
|
||||
|
||||
```bash
|
||||
sudo -i
|
||||
kubeadm token create --print-join-command
|
||||
```
|
||||
|
||||
可获取kubeadm join 命令及参数，如下所示
|
||||
|
||||
```bash
|
||||
kubeadm join apiserver.apple:6443 --token mpfjma.4vjjg8flqihor4vt --discovery-token-ca-cert-hash sha256:6f7a8e40a810323672de5eee6f4d19aa2dbdb38411845a1bf5dd63485c43d303
|
||||
```
|
||||
|
||||
|
||||
|
||||
### 初始化worker
|
||||
|
||||
**针对所有的 worker 节点执行**
|
||||
|
||||
```bash
|
||||
sudo -i
|
||||
echo "x.x.x.x apiserver.apple" >> /etc/hosts
|
||||
kubeadm join apiserver.apple:6443 --token mpfjma.4vjjg8flqihor4vt --discovery-token-ca-cert-hash sha256:6f7a8e40a810323672de5eee6f4d19aa2dbdb38411845a1bf5dd63485c43d303
|
||||
```
|
||||
|
||||
> * 将 x.x.x.x 替换为 loader balancer 的实际 ip
|
||||
>
|
||||
> * 将 kubeadm join 命令后的参数替换为上一个步骤中实际从 apple-master-a-1 节点获得的参数
|
||||
|
||||
|
||||
|
||||
### 检查 apiserver初始化结果
|
||||
|
||||
在第一个master节点 apple-master-a-1 上执行
|
||||
|
||||
```bash
|
||||
sudo -i
|
||||
su - gitlab-runner
|
||||
kubectl get nodes
|
||||
```
|
||||
|
||||
|
||||
|
||||
## 移除 worker 节点
|
||||
|
||||
> 正常情况下，您无需移除 worker 节点
|
||||
|
||||
在准备移除的 worker 节点上执行
|
||||
|
||||
```bash
|
||||
sudo -i
|
||||
kubeadm reset
|
||||
```
|
||||
|
||||
在第一个 master 节点 apple-master-a-1 上执行
|
||||
|
||||
```bash
|
||||
sudo -i
|
||||
su - gitlab-runner
|
||||
kubectl delete node apple-worker-x-x
|
||||
```
|
||||
|
||||
> * 将 apple-worker-x-x 替换为要移除的 worker 节点的名字
|
||||
> * worker 节点的名字可以通过在第一个 master 节点 apple-master-a-1 上执行 kubectl get nodes 命令获得
|
||||
|
||||
|
||||
|
||||
## 安装 Ingress Controller
|
||||
|
||||
> ingress官方文档：https://kubernetes.io/docs/concepts/services-networking/ingress/
|
||||
>
|
||||
> Ingress Controllers官网介绍：[https://kubernetes.io/docs/concepts/services-networking/ingress-controllers/](https://kubernetes.io/docs/concepts/services-networking/ingress-controllers/)
|
||||
>
|
||||
> 本文中使用如下部署方式：https://kubernetes.github.io/ingress-nginx/deploy/baremetal/#using-a-self-provisioned-edge
|
||||
>
|
||||
> kubernetes支持多种Ingress Controllers，本文推荐使用
|
||||
>
|
||||
> https://github.com/nginxinc/kubernetes-ingress
|
||||
|
||||
|
||||
|
||||
### 在 apple-master-a-1 上执行
|
||||
|
||||
```bash
|
||||
su - gitlab-runner
|
||||
kubectl apply -f https://raw.githubusercontent.com/eip-work/eip-monitor-repository/master/dashboard/nginx-ingress.yaml
|
||||
```
|
||||
|
||||
|
||||
|
||||
### 在IaaS云控制台完成如下配置（**公网ELB**）
|
||||
|
||||
创建负载均衡 ELB：
|
||||
|
||||
监听器 1：80 / TCP， SOURCE_ADDRESS 会话保持
|
||||
|
||||
服务器资源池 1： apple-worker-x-x 的所有节点的 80端口
|
||||
|
||||
监听器 2：443 / TCP， SOURCE_ADDRESS 会话保持
|
||||
|
||||
服务器资源池 2： apple-worker-x-x 的所有节点的443端口
|
||||
|
||||
假设刚创建的负载均衡 ELB 的 IP 地址为： z.z.z.z
|
||||
|
||||
|
||||
|
||||
### 配置域名解析
|
||||
|
||||
将域名 *.apple.yourdomain.com 解析到地址负载均衡服务器 的 IP 地址 z.z.z.z
|
||||
|
||||
|
||||
|
||||
### 验证配置
|
||||
|
||||
在浏览器访问 a.apple.yourdomain.com，将得到 404 NotFound 错误页面
|
||||
|
||||
Reference in New Issue
Block a user