kubernetes实战(二十五)：kubeadm安装高可用k8sv1.13.x

1、系统环境

站在用户的角度思考问题，与客户深入沟通，找到兖州网站设计与兖州网站推广的解决方案，凭借多年的经验，让设计与互联网技术结合，创造个性化、用户体验好的作品，建站类型包括：成都网站设计、网站制作、外贸营销网站建设、企业官网、英文网站、手机端网站、网站推广、域名注册、网页空间、企业邮箱。业务覆盖兖州地区。

　使用kubeadm安装高可用k8s v.13.x较为简单，相比以往的版本省去了很多步骤。

　kubeadm安装高可用k8s v.11 和 v1.12 点我

　主机信息

　主机配置

[root@k8s-master01 ~]# uname -r4.18.9-1.el7.elrepo.x86_64
[root@k8s-master01 ~]# uname -a
Linux k8s-master01 4.18.9-1.el7.elrepo.x86_64 #1 SMP Thu Sep 20 09:04:54 EDT 2018 x86_64 x86_64 x86_64 GNU/Linux
[root@k8s-master01 ~]# hostnamek8s-master01
[root@k8s-master01 ~]# free -g
              total        used        free      shared  buff/cache   available
Mem:              3           1           0           0           2           2Swap:             0           0           0[root@k8s-master01 ~]# cat /proc/cpuinfo | grep process
processor    : 0processor    : 1processor    : 2processor    : 3[root@k8s-master01 ~]# cat /etc/redhat-release 
CentOS Linux release 7.5.1804 (Core)

　Docker和k8s版本

[root@k8s-master01 ~]# docker version
Client:
 Version:      17.09.1-ce
 API version:  1.32
 Go version:   go1.8.3
 Git commit:   19e2cf6
 Built:        Thu Dec  7 22:23:40 2017
 OS/Arch:      linux/amd64

Server:
 Version:      17.09.1-ce
 API version:  1.32 (minimum version 1.12)
 Go version:   go1.8.3
 Git commit:   19e2cf6
 Built:        Thu Dec  7 22:25:03 2017
 OS/Arch:      linux/amd64
 Experimental: false[root@k8s-master01 ~]# kubectl version
Client Version: version.Info{Major:"1", Minor:"13", GitVersion:"v1.13.2", GitCommit:"cff46ab41ff0bb44d8584413b598ad8360ec1def", GitTreeState:"clean", BuildDate:"2019-01-10T23:35:51Z", GoVersion:"go1.11.4", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"13", GitVersion:"v1.13.2", GitCommit:"cff46ab41ff0bb44d8584413b598ad8360ec1def", GitTreeState:"clean", BuildDate:"2019-01-10T23:28:14Z", GoVersion:"go1.11.4", Compiler:"gc", Platform:"linux/amd64"}

2、配置SSH互信

　所有节点配置hosts：

[root@k8s-master01 ~]# cat /etc/hosts

192.168.20.20 k8s-master01
192.168.20.21 k8s-master02
192.168.20.22 k8s-master03
192.168.20.10 k8s-master-lb
192.168.20.30 k8s-node01
192.168.20.31 k8s-node02

　在k8s-master01上执行：

[root@k8s-master01 ~]# ssh-keygen -t rsaGenerating public/private rsa key pair.
Enter file in which to save the key (/root/.ssh/id_rsa): 
Created directory '/root/.ssh'.
Enter passphrase (empty for no passphrase): 
Enter same passphrase again: 
Your identification has been saved in /root/.ssh/id_rsa.
Your public key has been saved in /root/.ssh/id_rsa.pub.
The key fingerprint is:
SHA256:TE0eRfhGNRXL3btmmMRq+awUTkR4RnWrMf6Q5oJaTn0 root@k8s-master01
The key's randomart image is:
+---[RSA 2048]----+
|          =*+oo+o|
|         =o+. o.=|
|        . =+ o +o|
|       o  . = = .|
|        S  + O . |
|          = B = .|
|         + O E = |
|        = o = o  |
|       . . ..o   |
+----[SHA256]-----+

for i in k8s-master01 k8s-master02 k8s-master03 k8s-node01 k8s-node02;do ssh-copy-id -i .ssh/id_rsa.pub $i;done

　所有节点关闭防火墙和selinux

[root@k8s-master01 ~]# systemctl disable --now firewalld NetworkManager
Removed symlink /etc/systemd/system/multi-user.target.wants/NetworkManager.service.
Removed symlink /etc/systemd/system/multi-user.target.wants/firewalld.service.
Removed symlink /etc/systemd/system/dbus-org.freedesktop.NetworkManager.service.
Removed symlink /etc/systemd/system/dbus-org.freedesktop.nm-dispatcher.service.
Removed symlink /etc/systemd/system/dbus-org.fedoraproject.FirewallD1.service.
[root@k8s-master01 ~]# setenforce 0
[root@k8s-master01 ~]# sed -ri '/^[^#]*SELINUX=/s#=.+$#=disabled#' /etc/selinux/config

　所有节点关闭DNSmasq（如开启）

systemctl disable --now dnsmasq

　所有节点关闭swap

[root@k8s-master01 ~]# swapoff -a && sysctl -w vm.swappiness=0
vm.swappiness = 0
[root@k8s-master01 ~]# sed -ri '/^[^#]*swap/s@^@#@' /etc/fstab

　所有节点升级系统

yum install epel-release -y
yum install wget git  jq psmisc vim -y
yum update -y --exclude=kernel*

　所有节点同步时间

ln -sf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime
echo 'Asia/Shanghai' >/etc/timezone
ntpdate time2.aliyun.com
# 加入到crontab

　所有节点limit配置

ulimit -SHn 65535

　master01下载安装文件

[root@k8s-master01 ~]# git clone https://github.com/dotbalo/k8s-ha-install.git -b v1.13.x

　所有节点创建repo

cd /etc/yum.repos.d
mkdir bak
mv *.repo bak/
cp /root/k8s-ha-install/repo/* .

　所有节点升级系统并重启

yum install wget git  jq psmisc -yyum update -y --exclude=kernel* && reboot

　内核升级

　所有节点

[root@k8s-master01 ~]# rpm --import https://www.elrepo.org/RPM-GPG-KEY-elrepo.org[root@k8s-master01 ~]# rpm -Uvh http://www.elrepo.org/elrepo-release-7.0-2.el7.elrepo.noarch.rpmRetrieving http://www.elrepo.org/elrepo-release-7.0-2.el7.elrepo.noarch.rpmRetrieving http://elrepo.org/elrepo-release-7.0-3.el7.elrepo.noarch.rpmPreparing...                          ################################# [100%]
Updating / installing...   1:elrepo-release-7.0-3.el7.elrepo  ################################# [100%]

　master01下载内核文件

wget http://mirror.rc.usf.edu/compute_lock/elrepo/kernel/el7/x86_64/RPMS/kernel-ml-devel-4.18.9-1.el7.elrepo.x86_64.rpmwget http://mirror.rc.usf.edu/compute_lock/elrepo/kernel/el7/x86_64/RPMS/kernel-devel-4.18.9-1.el7.elrepo.x86_64.rpm

　复制到其他节点

[root@k8s-master01 ~]# for i in k8s-master01 k8s-master02 k8s-master03 k8s-node01 k8s-node02;do scp kernel-ml-4.18.9-1.el7.elrepo.x86_64.rpm kernel-ml-devel-4.18.9-1.el7.elrepo.x86_64.rpm $i:/root/ ; donekernel-ml-4.18.9-1.el7.elrepo.x86_64.rpm                                               100%   45MB 147.2MB/s   00:00    kernel-ml-devel-4.18.9-1.el7.elrepo.x86_64.rpm                                         100%   12MB 149.1MB/s   00:00    kernel-ml-4.18.9-1.el7.elrepo.x86_64.rpm                                               100%   45MB  22.6MB/s   00:02    kernel-ml-devel-4.18.9-1.el7.elrepo.x86_64.rpm                                         100%   12MB  20.8MB/s   00:00    kernel-ml-4.18.9-1.el7.elrepo.x86_64.rpm                                               100%   45MB  15.1MB/s   00:03    kernel-ml-devel-4.18.9-1.el7.elrepo.x86_64.rpm                                         100%   12MB  11.9MB/s   00:01    kernel-ml-4.18.9-1.el7.elrepo.x86_64.rpm                                               100%   45MB  45.1MB/s   00:01    kernel-ml-devel-4.18.9-1.el7.elrepo.x86_64.rpm                                         100%   12MB  27.4MB/s   00:00    kernel-ml-4.18.9-1.el7.elrepo.x86_64.rpm                                               100%   45MB  45.1MB/s   00:01    kernel-ml-devel-4.18.9-1.el7.elrepo.x86_64.rpm                                         100%   12MB  30.0MB/s   00:00

　所有节点安装内核

yum localinstall -y kernel-ml*

　所有节点修改内核启动顺序

grub2-set-default  0 && grub2-mkconfig -o /etc/grub2.cfg

grubby --args="user_namespace.enable=1" --update-kernel="$(grubby --default-kernel)"

　所有节点重启

reboot

　确认所有节点为新内核

[root@k8s-master01 ~]# uname -r4.18.9-1.el7.elrepo.x86_64

　确认能否加载nf_conntrack_ipv4，ipvs依赖此模块，需要确认能正常加载

[root@k8s-master02 ~]# modprobe nf_conntrack_ipv4
[root@k8s-master02 ~]# lsmod | grep nf
nf_conntrack_ipv4      16384  0 nf_defrag_ipv4         16384  1 nf_conntrack_ipv4
nf_conntrack          135168  1 nf_conntrack_ipv4
libcrc32c              16384  2 nf_conntrack,xfs

　所有节点安装ipvsadm

yum install ipvsadm ipset sysstat conntrack libseccomp -y

　所有节点设置开启自动加载的模块

[root@k8s-master01 ~]# cat /etc/modules-load.d/ipvs.conf 
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

[root@k8s-master01 ~]# systemctl disable --now systemd-modules-load.service

[root@k8s-master01 ~]# lsmod |grep ip_vs
ip_vs_ftp              16384  0 nf_nat                 32768  1 ip_vs_ftp
ip_vs_sed              16384  0 ip_vs_nq               16384  0 ip_vs_fo               16384  0 ip_vs_sh               16384  0 ip_vs_dh               16384  0 ip_vs_lblcr            16384  0 ip_vs_lblc             16384  0 ip_vs_wrr              16384  0 ip_vs_rr               16384  0 ip_vs_wlc              16384  0 ip_vs_lc               16384  0 ip_vs                 151552  24 ip_vs_wlc,ip_vs_rr,ip_vs_dh,ip_vs_lblcr,ip_vs_sh,ip_vs_fo,ip_vs_nq,ip_vs_lblc,ip_vs_wrrip_vs_lc,ip_vs_sed,ip_vs_ftp
nf_conntrack          135168  3 nf_conntrack_ipv4,nf_nat,ip_vs
libcrc32c              16384  4 nf_conntrack,nf_nat,xfs,ip_vs

　所有节点配置k8s内核

cat < /etc/sysctl.d/k8s.conf
net.ipv4.ip_forward = 1net.bridge.bridge-nf-call-ip6tables = 1net.bridge.bridge-nf-call-iptables = 1fs.may_detach_mounts = 1vm.overcommit_memory=1vm.panic_on_oom=0fs.inotify.max_user_watches=89100fs.file-max=52706963fs.nr_open=52706963net.netfilter.nf_conntrack_max=2310720EOF

sysctl --system

3、k8s服务安装

　所有节点安装docker-ce

yum -y install docker-ce-17.09.1.ce-1.el7.centos

　所有节点安装集群组件

yum install -y kubelet-1.13.2-0.x86_64 kubeadm-1.13.2-0.x86_64 kubectl-1.13.2-0.x86_64

　所有节点启动docker和kubelet

systemctl enable docker && systemctl start docker

[root@k8s-master01 ~]# DOCKER_CGROUPS=$(docker info | grep 'Cgroup' | cut -d' ' -f3)
[root@k8s-master01 ~]# echo $DOCKER_CGROUPS
cgroupfs
[root@k8s-master01 ~]# cat >/etc/sysconfig/kubelet< KUBELET_EXTRA_ARGS="--cgroup-driver=$DOCKER_CGROUPS --pod-infra-container-image=registry.cn-hangzhou.aliyuncs.com/google_containers/pause-amd64:3.1"> EOF
[root@k8s-master01 ~]# 
[root@k8s-master01 ~]# systemctl daemon-reload
[root@k8s-master01 ~]# systemctl enable kubelet && systemctl start kubelet
Created symlink from /etc/systemd/system/multi-user.target.wants/kubelet.service to /etc/systemd/system/kubelet.service.
[root@k8s-master01 ~]#

　注意此时如果kubelet无法启动不用管

　在所有master节点安装并启动keepalived及docker-compose

yum install -y keepalived
systemctl enable keepalived && systemctl restart keepalived

#安装docker-composeyum install -y docker-compose

4、master01节点安装

　以下操作在master01节点

　创建配置文件

　修改对应的配置信息，注意nm-bond修改为服务器对应网卡名称

[root@k8s-master01 k8s-ha-install]# ./create-config.sh create kubeadm-config.yaml files success. config/k8s-master01/kubeadm-config.yaml
create kubeadm-config.yaml files success. config/k8s-master02/kubeadm-config.yaml
create kubeadm-config.yaml files success. config/k8s-master03/kubeadm-config.yaml
create keepalived files success. config/k8s-master01/keepalived/create keepalived files success. config/k8s-master02/keepalived/create keepalived files success. config/k8s-master03/keepalived/create nginx-lb files success. config/k8s-master01/nginx-lb/create nginx-lb files success. config/k8s-master02/nginx-lb/create nginx-lb files success. config/k8s-master03/nginx-lb/create calico.yaml file success. calico/calico.yaml
[root@k8s-master01 k8s-ha-install]# pwd/root/k8s-ha-instal

　分发文件

[root@k8s-master01 k8s-ha-install]# export HOST1=k8s-master01
[root@k8s-master01 k8s-ha-install]# export HOST2=k8s-master02
[root@k8s-master01 k8s-ha-install]# export HOST3=k8s-master03
[root@k8s-master01 k8s-ha-install]# scp -r config/$HOST1/kubeadm-config.yaml $HOST1:/root/kubeadm-config.yaml                                                                    100%  993     1.9MB/s   00:00    [root@k8s-master01 k8s-ha-install]# scp -r config/$HOST2/kubeadm-config.yaml $HOST2:/root/kubeadm-config.yaml                                                                    100% 1071    63.8KB/s   00:00    [root@k8s-master01 k8s-ha-install]# scp -r config/$HOST3/kubeadm-config.yaml $HOST3:/root/kubeadm-config.yaml                                                                    100% 1112    27.6KB/s   00:00    [root@k8s-master01 k8s-ha-install]# scp -r config/$HOST1/keepalived/* $HOST1:/etc/keepalived/
check_apiserver.sh                                                                     100%  471    36.4KB/s   00:00    
keepalived.conf                                                                        100%  558    69.9KB/s   00:00    
You have new mail in /var/spool/mail/root
[root@k8s-master01 k8s-ha-install]# scp -r config/$HOST2/keepalived/* $HOST2:/etc/keepalived/

check_apiserver.sh                                                                     100%  471    10.8KB/s   00:00    
keepalived.conf                                                                        100%  558   275.5KB/s   00:00    
[root@k8s-master01 k8s-ha-install]# scp -r config/$HOST3/keepalived/* $HOST3:/etc/keepalived/
check_apiserver.sh                                                                     100%  471    12.7KB/s   00:00    
keepalived.conf                                                                        100%  558     1.1MB/s   00:00    
[root@k8s-master01 k8s-ha-install]# scp -r config/$HOST1/nginx-lb $HOST1:/root/
docker-compose.yaml                                                                    100%  213   478.6KB/s   00:00    
nginx-lb.conf                                                                          100% 1036     2.6MB/s   00:00    
[root@k8s-master01 k8s-ha-install]# scp -r config/$HOST2/nginx-lb $HOST2:/root/

docker-compose.yaml                                                                    100%  213    12.5KB/s   00:00    
nginx-lb.conf                                                                          100% 1036    35.5KB/s   00:00    
[root@k8s-master01 k8s-ha-install]# scp -r config/$HOST3/nginx-lb $HOST3:/root/
docker-compose.yaml                                                                    100%  213    20.5KB/s   00:00    
nginx-lb.conf                                                                          100% 1036    94.3KB/s   00:00

　所有master节点启动nginx

启动nginx-lb，修改nginx-lb下的nginx配置文件proxy_connect_timeout 60s;proxy_timeout 10m;
cd 
docker-compose --file=/root/nginx-lb/docker-compose.yaml up -d
docker-compose --file=/root/nginx-lb/docker-compose.yaml ps

　重启keepalived

systemctl restart keepalived

　提前下载镜像

kubeadm config images pull --config /root/kubeadm-config.yaml

　集群初始化

kubeadm init --config /root/kubeadm-config.yaml

....
kubeadm join k8s-master-lb:16443 --token cxwr3f.2knnb1gj83ztdg9l --discovery-token-ca-cert-hash sha256:41718412b5d2ccdc8b7326fd440360bf186a21dac4a0769f460ca4bdaf5d2825
....

[root@k8s-master01 ~]# cat <> ~/.bashrc
export KUBECONFIG=/etc/kubernetes/admin.conf
EOF
[root@k8s-master01 ~]# source ~/.bashrc
[root@k8s-master01 ~]# kubectl get nodes
NAME           STATUS     ROLES    AGE     VERSION
k8s-master01   NotReady   master   2m11s   v1.13.2

　查看pods状态

[root@k8s-master01 ~]# kubectl get pods -n kube-system -o wide
NAME                                   READY   STATUS              RESTARTS   AGE     IP              NODE           NOMINATED NODE   READINESS GATES
coredns-89cc84847-2h7r6                0/1     ContainerCreating   0          3m12s             k8s-master01              coredns-89cc84847-fhwbr                0/1     ContainerCreating   0          3m12s             k8s-master01              etcd-k8s-master01                      1/1     Running             0          2m31s   192.168.20.20   k8s-master01              kube-apiserver-k8s-master01            1/1     Running             0          2m36s   192.168.20.20   k8s-master01              kube-controller-manager-k8s-master01   1/1     Running             0          2m39s   192.168.20.20   k8s-master01              kube-proxy-kb95s                       1/1     Running             0          3m12s   192.168.20.20   k8s-master01              kube-scheduler-k8s-master01            1/1     Running             0          2m46s   192.168.20.20   k8s-master01              

　此时CoreDNS状态为ContainerCreating，报错如下：

  Normal   Scheduled        2m51s                  default-scheduler      Successfully assigned kube-system/coredns-89cc84847-2h7r6 to k8s-master01
  Warning  NetworkNotReady  2m3s (x25 over 2m51s)  kubelet, k8s-master01  network is not ready: [runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:docker: network plugin is not ready: cni config uninitialized

　因为没有安装网络插件，暂时不用管

　安装calico

[root@k8s-master01 k8s-ha-install]# kubectl create -f calico/configmap/calico-config created
service/calico-typha created
deployment.apps/calico-typha created
poddisruptionbudget.policy/calico-typha created
daemonset.extensions/calico-node created
serviceaccount/calico-node created
customresourcedefinition.apiextensions.k8s.io/felixconfigurations.crd.projectcalico.org created
customresourcedefinition.apiextensions.k8s.io/bgppeers.crd.projectcalico.org created
customresourcedefinition.apiextensions.k8s.io/bgpconfigurations.crd.projectcalico.org created
customresourcedefinition.apiextensions.k8s.io/ippools.crd.projectcalico.org created
customresourcedefinition.apiextensions.k8s.io/hostendpoints.crd.projectcalico.org created
customresourcedefinition.apiextensions.k8s.io/clusterinformations.crd.projectcalico.org created
customresourcedefinition.apiextensions.k8s.io/globalnetworkpolicies.crd.projectcalico.org created
customresourcedefinition.apiextensions.k8s.io/globalnetworksets.crd.projectcalico.org created
customresourcedefinition.apiextensions.k8s.io/networkpolicies.crd.projectcalico.org created
clusterrole.rbac.authorization.k8s.io/calico-node created
clusterrolebinding.rbac.authorization.k8s.io/calico-node created

　再次查看

[root@k8s-master01 k8s-ha-install]# kubectl get po -n kube-system
NAME                                   READY   STATUS    RESTARTS   AGE
calico-node-tp2dz                      2/2     Running   0          42s
coredns-89cc84847-2djpl                1/1     Running   0          66s
coredns-89cc84847-vt6zq                1/1     Running   0          66s
etcd-k8s-master01                      1/1     Running   0          27s
kube-apiserver-k8s-master01            1/1     Running   0          16s
kube-controller-manager-k8s-master01   1/1     Running   0          34s
kube-proxy-x497d                       1/1     Running   0          66s
kube-scheduler-k8s-master01            1/1     Running   0          17s

5、高可用配置

　复制证书

USER=root
CONTROL_PLANE_IPS="k8s-master02 k8s-master03"for host in $CONTROL_PLANE_IPS; do
    ssh "${USER}"@$host "mkdir -p /etc/kubernetes/pki/etcd"
    scp /etc/kubernetes/pki/ca.crt "${USER}"@$host:/etc/kubernetes/pki/ca.crt    scp /etc/kubernetes/pki/ca.key "${USER}"@$host:/etc/kubernetes/pki/ca.key    scp /etc/kubernetes/pki/sa.key "${USER}"@$host:/etc/kubernetes/pki/sa.key    scp /etc/kubernetes/pki/sa.pub "${USER}"@$host:/etc/kubernetes/pki/sa.pub    scp /etc/kubernetes/pki/front-proxy-ca.crt "${USER}"@$host:/etc/kubernetes/pki/front-proxy-ca.crt    scp /etc/kubernetes/pki/front-proxy-ca.key "${USER}"@$host:/etc/kubernetes/pki/front-proxy-ca.key    scp /etc/kubernetes/pki/etcd/ca.crt "${USER}"@$host:/etc/kubernetes/pki/etcd/ca.crt    scp /etc/kubernetes/pki/etcd/ca.key "${USER}"@$host:/etc/kubernetes/pki/etcd/ca.key    scp /etc/kubernetes/admin.conf "${USER}"@$host:/etc/kubernetes/admin.confdone

　以下操作在master02执行

　提前下载镜像

kubeadm config images pull --config /root/kubeadm-config.yaml

　master02加入集群，与node节点相差的参数就是--experimental-control-plane

kubeadm join k8s-master-lb:16443 --token cxwr3f.2knnb1gj83ztdg9l --discovery-token-ca-cert-hash sha256:41718412b5d2ccdc8b7326fd440360bf186a21dac4a0769f460ca4bdaf5d2825 --experimental-control-plane

......

This node has joined the cluster and a new control plane instance was created:* Certificate signing request was sent to apiserver and approval was received.* The Kubelet was informed of the new secure connection details.* Master label and taint were applied to the new node.* The Kubernetes control plane instances scaled up.* A new etcd member was added to the local/stacked etcd cluster.

To start administering your cluster from this node, 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

Run 'kubectl get nodes' to see this node join the cluster.

　master01查看状态

[root@k8s-master01 k8s-ha-install]# kubectl get no
NAME           STATUS   ROLES    AGE     VERSION
k8s-master01   Ready    master   15m     v1.13.2k8s-master02   Ready    master   9m55s   v1.13.2

　其他master节点类似

　查看最终master状态

[root@k8s-master01 ~]# ipvsadm -lnIP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  10.96.0.1:443 rr  -> 192.168.20.20:6443           Masq    1      4          0         
  -> 192.168.20.21:6443           Masq    1      0          0         
  -> 192.168.20.22:6443           Masq    1      0          0         TCP  10.96.0.10:53 rr  -> 172.168.0.10:53              Masq    1      0          0         
  -> 172.168.0.11:53              Masq    1      0          0         TCP  10.102.221.48:5473 rr
UDP  10.96.0.10:53 rr  -> 172.168.0.10:53              Masq    1      0          0         
  -> 172.168.0.11:53              Masq    1      0          0         [root@k8s-master01 ~]# kubectl get po -n kube-system
NAME                                   READY   STATUS    RESTARTS   AGE
calico-node-49dwr                      2/2     Running   0          26m
calico-node-kz2d4                      2/2     Running   0          22m
calico-node-zwnmq                      2/2     Running   0          4m6s
coredns-89cc84847-dgxlw                1/1     Running   0          27m
coredns-89cc84847-n77x6                1/1     Running   0          27m
etcd-k8s-master01                      1/1     Running   0          27m
etcd-k8s-master02                      1/1     Running   0          22m
etcd-k8s-master03                      1/1     Running   0          4m5s
kube-apiserver-k8s-master01            1/1     Running   0          27m
kube-apiserver-k8s-master02            1/1     Running   0          22m
kube-apiserver-k8s-master03            1/1     Running   3          4m6s
kube-controller-manager-k8s-master01   1/1     Running   1          27m
kube-controller-manager-k8s-master02   1/1     Running   0          22m
kube-controller-manager-k8s-master03   1/1     Running   0          4m6s
kube-proxy-f9qc5                       1/1     Running   0          27m
kube-proxy-k55bg                       1/1     Running   0          22m
kube-proxy-kbg9c                       1/1     Running   0          4m6s
kube-scheduler-k8s-master01            1/1     Running   1          27m
kube-scheduler-k8s-master02            1/1     Running   0          22m
kube-scheduler-k8s-master03            1/1     Running   0          4m6s
[root@k8s-master01 ~]# kubectl get no
NAME           STATUS   ROLES    AGE     VERSION
k8s-master01   Ready    master   28m     v1.13.2k8s-master02   Ready    master   22m     v1.13.2k8s-master03   Ready    master   4m16s   v1.13.2

[root@k8s-master01 ~]# kubectl get csr
NAME                                                   AGE     REQUESTOR                  CONDITION
csr-6mqbv                                              28m     system:node:k8s-master01   Approved,Issued
node-csr-GPLcR1G4Nchf-zuB5DaTWncoluMuENUfKvWKs0j2GdQ   23m     system:bootstrap:9zp70m    Approved,Issued
node-csr-cxAxrkllyidkBuZ8fck6fwq-ht1_u6s0snbDErM8bIs   4m51s   system:bootstrap:9zp70m    Approved,Issued

　在所有master节点上允许hpa通过接口采集数据

vi /etc/kubernetes/manifests/kube-controller-manager.yaml
    - --horizontal-pod-autoscaler-use-rest-clients=false

　在所有master上允许istio的自动注入，修改/etc/kubernetes/manifests/kube-apiserver.yaml

vi /etc/kubernetes/manifests/kube-apiserver.yaml
    - --enable-admission-plugins=NamespaceLifecycle,LimitRanger,ServiceAccount,DefaultStorageClass,DefaultTolerationSeconds,MutatingAdmissionWebhook,ValidatingAdmissionWebhook,ResourceQuota

6、node节点加入集群

kubeadm join 192.168.20.10:16443 --token ll4usb.qmplnofiv7z1j0an  --discovery-token-ca-cert-hash sha256:e88a29f62ab77a59bf88578abadbcd37e89455515f6ecf3ca371656dc65b1d6e

......
[kubelet-start] Activating the kubelet service
[tlsbootstrap] Waiting for the kubelet to perform the TLS Bootstrap...
[patchnode] Uploading the CRI Socket information "/var/run/dockershim.sock" to the Node API object "k8s-node02" as an annotation

This node has joined the cluster:* Certificate signing request was sent to apiserver and a response was received.* The Kubelet was informed of the new secure connection details.

Run 'kubectl get nodes' on the master to see this node join the cluster.

　master节点查看

[root@k8s-master01 k8s-ha-install]# kubectl get po -n kube-system
NAME                                   READY   STATUS    RESTARTS   AGE
calico-node-49dwr                      2/2     Running   0          13h
calico-node-9nmhb                      2/2     Running   0          11m
calico-node-k5nmt                      2/2     Running   0          11m
calico-node-kz2d4                      2/2     Running   0          13h
calico-node-zwnmq                      2/2     Running   0          13h
coredns-89cc84847-dgxlw                1/1     Running   0          13h
coredns-89cc84847-n77x6                1/1     Running   0          13h
etcd-k8s-master01                      1/1     Running   0          13h
etcd-k8s-master02                      1/1     Running   0          13h
etcd-k8s-master03                      1/1     Running   0          13h
kube-apiserver-k8s-master01            1/1     Running   0          18m
kube-apiserver-k8s-master02            1/1     Running   0          17m
kube-apiserver-k8s-master03            1/1     Running   0          16m
kube-controller-manager-k8s-master01   1/1     Running   0          19m
kube-controller-manager-k8s-master02   1/1     Running   1          19m
kube-controller-manager-k8s-master03   1/1     Running   0          19m
kube-proxy-cl2zv                       1/1     Running   0          11m
kube-proxy-f9qc5                       1/1     Running   0          13h
kube-proxy-hkcq5                       1/1     Running   0          11m
kube-proxy-k55bg                       1/1     Running   0          13h
kube-proxy-kbg9c                       1/1     Running   0          13h
kube-scheduler-k8s-master01            1/1     Running   1          13h
kube-scheduler-k8s-master02            1/1     Running   0          13h
kube-scheduler-k8s-master03            1/1     Running   0          13h
You have new mail in /var/spool/mail/root
[root@k8s-master01 k8s-ha-install]# kubectl get no
NAME           STATUS   ROLES    AGE   VERSION
k8s-master01   Ready    master   13h   v1.13.2k8s-master02   Ready    master   13h   v1.13.2k8s-master03   Ready    master   13h   v1.13.2k8s-node01     Ready       11m   v1.13.2k8s-node02     Ready       11m   v1.13.2

7、其他组件安装

　部署metrics server 0.3.1/1.8+安装

[root@k8s-master01 k8s-ha-install]# kubectl create -f metrics-server/clusterrole.rbac.authorization.k8s.io/system:aggregated-metrics-reader created
clusterrolebinding.rbac.authorization.k8s.io/metrics-server:system:auth-delegator created
rolebinding.rbac.authorization.k8s.io/metrics-server-auth-reader created
apiservice.apiregistration.k8s.io/v1beta1.metrics.k8s.io created
serviceaccount/metrics-server created
deployment.extensions/metrics-server created
service/metrics-server created
clusterrole.rbac.authorization.k8s.io/system:metrics-server&nbs            
            
                        

            名称栏目：kubernetes实战(二十五)：kubeadm安装高可用k8sv1.13.x            

            浏览地址：http://shouzuofang.com/article/gsdgso.html