k8s-1.21高可用集群搭建_kubernetes 1.21 高可用集群搭建

前言：
1）本次搭建使用了docker作为引擎
抛弃docker的文档见：《k8s1.20搭建抛弃docker使用containerd》
2）如果是离线安装，镜像准备请看文章《k8s-1.21 安装——镜像准备》

环境说明

系统：centos 7.9

• 规划如下：

• 镜像准备

如离线部署，需要提前下载镜像，见文档 《k8s-1.21 安装——镜像准备》

• hostname
  按节点名修改每台服务器的hostanme。

• hosts
  修改 /etc/hosts如下；

10.10.239.201  crust-m01
10.10.239.202  crust-m02
10.10.239.203  crust-m03
10.10.239.204  crust-n01
10.10.239.205  crust-n02
10.10.239.206  crust-n03
1
2
3
4
5
6

1. 环境准备

1.1 关闭无用服务

• 关闭selinux

# sed -i 's#SELINUX=enforcing#SELINUX=disabled#g'  /etc/sysconfig/selinux
# sed -i 's#SELINUX=enforcing#SELINUX=disabled#g'  /etc/selinux/config
1
2

• 关闭交换分区

# swapoff -a
1

修改/etc/fstab文件，注释掉 swap一行。

• 关闭其他服务

有就关闭，没有报错就报错没关系。

# systemctl disable auditd
# systemctl disable postfix
# systemctl disable irqbalance
# systemctl disable remote-fs
# systemctl disable tuned
# systemctl disable rhel-configure
# systemctl disable firewalld
1
2
3
4
5
6
7

1.2 安装依赖

• 修改yum源

# mkdir /etc/yum.repos.d/bak && cp -rf /etc/yum.repos.d/*.repo /etc/yum.repos.d/bak
# wget -O /etc/yum.repos.d/CentOS-Base.repo http://mirrors.cloud.tencent.com/repo/centos7_base.repo
# wget -O /etc/yum.repos.d/epel.repo http://mirrors.cloud.tencent.com/repo/epel-7.repo
# yum clean all && yum makecache
# cat <<EOF > /etc/yum.repos.d/kubernetes.repo
 [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
1
2
3
4
5
6
7
8
9
10
11
12
13

• 安装依赖和升级

# yum -y install vim-enhanced wget curl net-tools conntrack-tools bind-utils socat ipvsadm ipset
# yum -y update
1
2

• 安装docker

自己做docker 1.19 离线部署包： 下载连接
执行部署包中的compose_install.sh脚本即可。

它会把docker-compose也一起安上，虽然没什么用，但是一个二进制文件对环境也没有影响。

1.3 服务器优化

• 内核优化

# cat >>/etc/sysctl.conf <<EOF
net.ipv4.ip_forward = 1
vm.swappiness = 0
net.bridge.bridge-nf-call-ip6tables = 1
net.bridge.bridge-nf-call-iptables = 1
net.ipv4.tcp_max_syn_backlog = 65536
net.core.netdev_max_backlog =  32768
net.core.somaxconn = 32768
net.core.wmem_default = 8388608
net.core.rmem_default = 8388608
net.core.rmem_max = 16777216
net.core.wmem_max = 16777216
net.ipv4.tcp_timestamps = 0
net.ipv4.tcp_synack_retries = 2
net.ipv4.tcp_syn_retries = 2
net.ipv4.tcp_tw_recycle = 1
net.ipv4.tcp_tw_reuse = 1
net.ipv4.tcp_mem = 94500000 915000000 927000000
net.ipv4.tcp_max_orphans = 3276800
net.ipv4.ip_local_port_range = 1024  65535
EOF

# sysctl -p
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

• 修改句柄数

设置临时生效

ulimit -n 65535
1

设置永久生效

# cat >>/etc/security/limits.conf <<EOF
* soft memlock unlimited
* hard memlock unlimited
* soft nofile 1048576
* hard nofile 1048576
* soft nproc 1048576
* hard nproc 1048576
EOF
1
2
3
4
5
6
7
8

全局配置（否则centos7永久设置不生效）
修改/etc/systemd/system.conf文件

DefaultLimitNOFILE=1048576
1

• 加载内核模块

# cat <<EOF>/etc/sysconfig/modules/ipvs.modules 
#!/bin/bash
ipvs_modules="ip_vs ip_vs_lc ip_vs_wlc ip_vs_rr ip_vs_wrr ip_vs_lblc ip_vs_lblcr ip_vs_dh ip_vs_sh ip_vs_fo ip_vs_nq ip_vs_sed ip_vs_ftp nf_conntrack_ipv4"
for kernel_module in \${ipvs_modules}; do
 /sbin/modinfo -F filename \${kernel_module} > /dev/null 2>&1
 if [ $? -eq 0 ]; then
 /sbin/modprobe \${kernel_module}
 fi
done
EOF

# chmod +x /etc/sysconfig/modules/ipvs.modules
# sh /etc/sysconfig/modules/ipvs.modules
1
2
3
4
5
6
7
8
9
10
11
12
13

2. master的高可用方案

说明：
使用keepalived+VIP 方案完成VIP的漂移。
使用HAProxy 完成负载。
另外：HAProxy和keepalive我们部署在每台master上，也可以单独部署。

2.1 安装以及配置haproxy

【三个master上操作】

说明：将api默认的6443接口由HAProxy通过 7443分发。因此后边可以看到，加入节点的时候使用的API接口用的是 VIP：7443

• 安装

# yum -y install haproxy
1

• 修改配置
  修改/etc/haproxy/haproxy.cfg文件

global
    #   /etc/sysconfig/syslog
    #
    #    local2.*                       /var/log/haproxy.log
    #
    log         127.0.0.1 local2

    chroot      /var/lib/haproxy
    pidfile     /var/run/haproxy.pid
    maxconn     4000
    user        haproxy
    group       haproxy
    daemon

defaults
    mode                    tcp
    log                     global
    retries                 3
    timeout connect         10s
    timeout client          1m
    timeout server          1m

frontend  kubernetes
        bind *:7443
        mode tcp
        default_backend kubernetes_master

backend kubernetes_master
    balance     roundrobin
    server master01 10.10.239.201:6443 check maxconn 2000
    server master02 10.10.239.202:6443 check maxconn 2000
    server master03 10.10.239.203:6443 check maxconn 2000
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

• 启动

# systemctl start haproxy
# systemctl enable haproxy
1
2

2.2 keepalived

• 安装y

# yum -y install keepalived
1

• 修改配置文件
  修改 /etc/keepalived/keepalived.conf，从节点按注释修改。

global_defs {
   router_id LVS_DEVEL
   vrrp_skip_check_adv_addr
   // vrrp_strict 
   vrrp_garp_interval 0
   vrrp_gna_interval 0
}

vrrp_instance VI_1 {
    state MASTER   #从节点为BACKUP
    interface eth0  #根据实际情况修改
    virtual_router_id 200
    priority 100   #master最高，根据情况修改  
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        10.10.239.200    #虚拟ip
    }
}
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

• 启动服务

# service keepalived start
# systemctl enable keepalive
1
2

3. k8s安装

3.1 安装kubeadm、kubelet、kubect

【所有节点上执行】

# yum install -y kubelet kubeadm kubectl
# systemctl enable kubelet
1
2

3.2 初始化集群

• 初始化文件
  【crust-01上操作】

apiVersion: kubeadm.k8s.io/v1beta2
kind: InitConfiguration
bootstrapTokens:
- groups:
  - system:bootstrappers:kubeadm:default-node-token
  token: abcdef.0123456789abcdef
  ttl: 24h0m0s
  usages:
  - signing
  - authentication
localAPIEndpoint:
  advertiseAddress: 10.10.239.201
  bindPort: 6443
nodeRegistration:
  criSocket: /var/run/dockershim.sock
  name: crust-m01
  taints:
  - effect: NoSchedule
    key: node-role.kubernetes.io/master
---
apiVersion: kubeadm.k8s.io/v1beta2
kind: ClusterConfiguration
clusterName: kubernetes
kubernetesVersion: v1.21.2
certificatesDir: /etc/kubernetes/pki
controllerManager: {}
controlPlaneEndpoint: "10.10.239.200:7443"
imageRepository: registry.aliyuncs.com/google_containers
apiServer:
  timeoutForControlPlane: 4m0s
  certSANs:
  - "crust-m01"
  - "crust-m02"
  - "crust-m03"
  - "10.10.239.200"
  - "10.10.239.201"
  - "10.10.239.202"
  - "10.10.239.203"
  - "127.0.0.1"
dns:
  type: CoreDNS
etcd:
  local:
    dataDir: /var/lib/etcd
networking:
  dnsDomain: cluster.local
  serviceSubnet: 10.1.0.0/16
  podSubnet: 10.244.0.0/16
scheduler: {}
---
apiVersion: kubeproxy.config.k8s.io/v1alpha1
kind: KubeProxyConfiguration
mode: "ipvs"
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

证书部分apiServer.certSANs，需加入能通过其访问API的地址或主机名（如master/Haproxy/vip等）
networking中service和pod的网段，后边网络插件你不管用 flannel或者calicol，都要和它保持一致。
imageRepository: 的值可以定义为阿里云也可以定义你自己的私有仓库。

• 初始化集群

# kubeadm init --config=init.yaml
# systemctl enable kubelet
1
2

• 输出信息

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/

You can now join any number of control-plane nodes by copying certificate authorities 
and service account keys on each node and then running the following as root:

  kubeadm join 10.10.239.200:7443 --token abcdef.0123456789abcdef \
        --discovery-token-ca-cert-hash sha256:b37367fb8a954bf72a8e54c2937c53cddfa09e2431dfbdbb30e0e54f2e8d6dc0 \
        --control-plane	  

Then you can join any number of worker nodes by running the following on each as root:

kubeadm join 10.10.239.200:7443 --token abcdef.0123456789abcdef \
        --discovery-token-ca-cert-hash sha256:b37367fb8a954bf72a8e54c2937c53cddfa09e2431dfbdbb30e0e54f2e8d6dc0

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

【说明】
以上输出了下边几条有用信息：

• 如何配置kubectl工具
• 如何接入其他master节点
• 如何接入node节点

• 配置kubectl工具

安装前边数据的信息执行即可

# mkdir -p $HOME/.kube
# cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
# chown $(id -u):$(id -g) $HOME/.kube/config
1
2
3

• kubectl命令补全
  见文档《kubectl命令补全》

3.3 master从节点安装

• 从节点拷贝证书
  【crust-01上操作】

# ssh 10.10.239.202 "mkdir -p /etc/kubernetes/pki/etcd"
# ssh 10.10.239.203 "mkdir -p /etc/kubernetes/pki/etcd"

# scp -r /etc/kubernetes/admin.conf 10.10.239.202:/etc/kubernetes/admin.conf
# scp -r /etc/kubernetes/admin.conf 10.10.239.203:/etc/kubernetes/admin.conf

# scp -r /etc/kubernetes/pki/{ca.*,sa.*,front*}  10.10.239.202:/etc/kubernetes/pki/
# scp -r /etc/kubernetes/pki/{ca.*,sa.*,front*}  10.10.239.203:/etc/kubernetes/pki/

# scp -r /etc/kubernetes/pki/etcd/ca.*  10.10.239.202:/etc/kubernetes/pki/etcd/ 
# scp -r /etc/kubernetes/pki/etcd/ca.*  10.10.239.203:/etc/kubernetes/pki/etcd/
1
2
3
4
5
6
7
8
9
10
11

• 接入master从节点
  【另外两个master上执行】、

初始化的时候输出了这条命令，我们复制过来执行即可。

  # kubeadm join 10.10.239.200:7443 --token abcdef.0123456789abcdef \
        --discovery-token-ca-cert-hash sha256:b37367fb8a954bf72a8e54c2937c53cddfa09e2431dfbdbb30e0e54f2e8d6dc0 \
        --control-plane	  
1
2
3

• 配置kubectl工具

# mkdir -p $HOME/.kube
# cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
# chown $(id -u):$(id -g) $HOME/.kube/config
1
2
3

3.4 网络插件

有很多选择，
一般小的器群我选择 flannel，见 《flannel网络的安装和删除》
大的集群选择calico，见《calico网络安装和删除》

3.5 node节点接入

【所有node节点上执行】、

初始化的时候输出了这条命令，我们复制过来执行即可。

  # kubeadm join 10.10.239.200:7443 --token abcdef.0123456789abcdef \
        --discovery-token-ca-cert-hash sha256:b37367fb8a954bf72a8e54c2937c53cddfa09e2431dfbdbb30e0e54f2e8d6dc0   
1
2

4. 查看结果

• 查看node

没有网络插件时，如果查看node节点，会显示NotReady。

[root@crust-m03 ~]# kubectl get node
NAME        STATUS   ROLES                  AGE     VERSION
crust-m01   Ready    control-plane,master   24h     v1.21.2
crust-m02   Ready    control-plane,master   23h     v1.21.2
crust-m03   Ready    control-plane,master   23h     v1.21.2
crust-n01   Ready    <none>                 3h22m   v1.21.2
crust-n02   Ready    <none>                 22m     v1.21.2
crust-n03   Ready    <none>                 17m     v1.21.2
1
2
3
4
5
6
7
8

• kube-system中的pod

没有接入网络插件的时候 coredns 会显示 Pending

[root@crust-m01 ~]# kubectl get pod -n kube-system
NAME                                READY   STATUS    RESTARTS   AGE
coredns-59d64cd4d4-8h9b7            1/1     Running   0          15h
coredns-59d64cd4d4-mpfln            1/1     Running   0          15h
etcd-crust-m01                      1/1     Running   2          24h
etcd-crust-m02                      1/1     Running   0          24h
etcd-crust-m03                      1/1     Running   0          24h
kube-apiserver-crust-m01            1/1     Running   2          24h
kube-apiserver-crust-m02            1/1     Running   0          24h
kube-apiserver-crust-m03            1/1     Running   0          24h
kube-controller-manager-crust-m01   1/1     Running   2          24h
kube-controller-manager-crust-m02   1/1     Running   0          24h
kube-controller-manager-crust-m03   1/1     Running   0          24h
kube-flannel-ds-amd64-8fw85         1/1     Running   0          31m
kube-flannel-ds-amd64-b6xts         1/1     Running   0          15h
kube-flannel-ds-amd64-gl542         1/1     Running   0          27m
kube-flannel-ds-amd64-krdzq         1/1     Running   0          3h31m
kube-flannel-ds-amd64-lqv9p         1/1     Running   0          15h
kube-flannel-ds-amd64-wndqk         1/1     Running   0          15h
kube-proxy-7rqmj                    1/1     Running   0          20m
kube-proxy-8s68r                    1/1     Running   0          3h31m
kube-proxy-hqvbs                    1/1     Running   0          31m
kube-proxy-srjq2                    1/1     Running   0          24h
kube-proxy-t6mvq                    1/1     Running   1          24h
kube-proxy-wbs9h                    1/1     Running   0          24h
kube-scheduler-crust-m01            1/1     Running   2          24h
kube-scheduler-crust-m02            1/1     Running   0          24h
kube-scheduler-crust-m03            1/1     Running   0          24h

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

5. 测试

启动一个nginx测试一下（镜像是我私有镜像，注意更换）

• 创建nginx.yml文件

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx
spec:
  replicas: 1
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: harbocto.xxx.com.cn/public/nginx
        imagePullPolicy: IfNotPresent
        ports:
        - containerPort: 80
---
apiVersion: v1
kind: Service
metadata:
  name: nginx
spec:
  type: NodePort
  ports:
  - port: 80
    targetPort: 80
    nodePort: 30200
  selector:
    app: nginx
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

• 创建nginx

[root@calico-master ~]# kubectl apply -f nginx.yml
deployment.apps/nginx created
service/nginx created
1
2
3

• 查看创建结果

[root@calico-master ~]# kubectl get pod
NAME                     READY   STATUS    RESTARTS   AGE
nginx-6764986cbd-ztdjf   1/1     Running   0          5s
[root@calico-master ~]# kubectl get service
NAME         TYPE        CLUSTER-IP     EXTERNAL-IP   PORT(S)        AGE
kubernetes   ClusterIP   10.1.0.1       <none>        443/TCP        2d2h
nginx        NodePort    10.1.157.245   <none>        80:30200/TCP   12s
1
2
3
4
5
6
7

• web上查看

6. storageclass 的安装

• 安装
  见文档 《K8S对象-StorageClass》
• FAQ
  《【FAQ】1.21创建PV报错》

---