基础环境

• 系统要求

1. intel 处理器 不低于8G内存 4核CPU
2. centos7.4 64位 最好Minimal安装

• 软件要求

1. intel 处理器 不低于8G内存 4核CPU
2. centos7.4 64位 最好Minimal安装

• 操作系统设置

1. systemctl stop firewalld
2. systemctl disable firewalld
3. iptables -F && iptables -Z && iptables -X
4. setenforce 0
5. swapoff -a
6. vi /etc/hosts

根据下面k8s 准备机器设置

192.168.1.32 master
192.168.1.32 etcd
192.168.1.33 node1
192.168.1.34 node2

docker 私有库安装

• docker安装

yum install docker -y

• 设置docker环境变量(镜像加速)

vi /etc/docker/deamon.json

{
 "registry-mirrors": ["https://registry.docker-cn.com","http://f2d6cb40.m.daocloud.io"],
 "insecure-registries": ["192.168.1.32:5000"]
}

• 启动docker

sytstemctl start docer

• 创建是私有库

docker run -d -p 5000:5000 --privileged=true -v /data/history:/data/registry registry --restart=always

• 推送

docker tag xxxx:1.0 192.168.1.32:5000/efei/xxxx:1.0
docker push 192.168.1.32:5000/efei/xxxx:1.0

Kubernets 安装

• 服务器准备

IP	用途
192.168.1.32	master
192.168.1.33	node
192.168.1.34	node

• 软件安装清单

IP	用途
192.168.1.32	kube-apiserver, kubecontroller-manager,kube-scheduler,flannel,kube-dns,docker,etcd
192.168.1.33	kubelet, proxy,docker,flannel
192.168.1.34	kubelet, proxy,docker,flannel

• 设置阿里云源

官网地址 : https://opsx.alibaba.com/

curl -o /etc/yum.repos.d/CentOS-Base.repo http://mirrors.aliyun.com/repo/Centos-7.repo

• 设置阿里kubernets源

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
setenforce 0

• 生效yum设置

yum clean
yum makecache

---

• 安装etcd

yum install etcd -y

• 配置etcd

vi /etc/etcd/etcd.conf
ETCD_NAME=default
ETCD_DATA_DIR="/var/lib/etcd/default.etcd"
ETCD_LISTEN_CLIENT_URLS="http://0.0.0.0:2379"
ETCD_ADVERTISE_CLIENT_URLS="http://192.168.1.32:2379"

• 启动etcd

systemctl start etcd

• 检查etcd

ss -anl|grep 2379
etcdctl cluster-health

---

k8s-Flannel 网络安装

• 设置网络地址分配范围

etcdctl set /k8s/network/config ‘{ “Network”: “172.16.0.0/16” }’

• 安装Flannel

yum install flannel -y

注：master、node机器都要安装

• 配置config

vi /etc/sysconfig/flanneld

FLANNEL_ETCD_ENDPOINTS="http://master:2379"
FLANNEL_ETCD_PREFIX="/k8s/network"

• 启动

systemctl start flanneld

• 注意事项

如果flannel在docker安装之后装的话需要执行下面操作

1. 启动flannel之前最好清空iptables，关闭 docker服务 及docker0网卡

iptables -X && iptables -F %% iptables -z
ip link delete docker0
rm -f /run/flannel/docker
rm -f /run/flannel/subnet.env
iptables -P FORWARD ACCEPT （可以放在安装完成之后再执行）

2. 重载系统配置并启动docker

systemctl daemon-reload
systemctl restart docker

3. 如果已经安装了k8s话 还要重启k8s(含master、node)

<重启请查看命令集>

---

• 安装master

yum install kubernetes-master -y
注：这个命令会安装 kubecontroller-manager,kube-scheduler 三个程序

• 配置apiserver

vi /etc/kubernets/apiserver

KUBE_ETCD_SERVERS="--etcd-servers=http://etcd:2379"
KUBE_SERVICE_ADDRESSES="--service-cluster-ip-range=10.244.0.0/16 --service-node-port-range=3000-65535"
KUBE_ADMISSION_CONTROL="--admission-control=NamespaceLifecycle,NamespaceExists,LimitRanger,ResourceQuota"

• 配置config

vi /etc/kubernets/config

KUBE_MASTER="--master=http://master:8080"

• 启动master

systemctl start kube-apiserver
systemctl start kube-scheduler
systemctl start kube-controller-manager

---

• 安装node

yum install kubernetes-node -y
这个命令会安装 kubelet, proxy,docker 三个程序

• 配置config

vi /etc/kubernets/config

KUBE_MASTER="--master=http://master:8080"

• 配置kuber

vi /etc/kubernets/kuber

KUBELET_ADDRESS="--address=0.0.0.0"
KUBELET_HOSTNAME="--hostname-override=node2"
KUBELET_API_SERVER="--api-servers=http://master:8080"
KUBELET_POD_INFRA_CONTAINER="--pod-infra-container-image=192.168.1.32:5000/rhel7/pod-infrastructure:latest"
KUBELET_ARGS="--cluster-dns=172.16.1.3 --cluster-domain=cluster.local."

注意:镜像要提前下载并推送到私有库中， dns可以安装dns再配置

• 启动node

systemctl start kubelet
systemctl start kube-proxy

---

k8s-dns 安装

为了容器内部能通过service访问，才需要安装dns,
否则非必须安装

• 安装 kube-dns-rc.yaml

配置文件需要修改注意地方

- --domain=cluster.local. 
- --kube-master-url = http://192.168.1.32:8080
- --cmd=nslookup kubernetes.default.svc.cluster.local. 127.0.0.1 >/dev/null
- --cmd=nslookup kubernetes.default.svc.cluster.local. 127.0.0.1:10053 >/dev/null

主要设置domain、master-url，目前不清楚doamain要不要加'.' , 我的猜想应该只要名字一致即可.

参考文件：kube-dns-rc.yaml

• 安装 kube-dns-svc.yaml

clusterIP: 10.244.2.2

参考文件：kube-dns-svc.yaml

• 测试dns

1. 安装 busybox.yaml

2. 进入busbox 容器

kubectl exec -i -t busybox sh

3. 使用 nslookup 测试

nslookup kubernetes

也可以直接： kubectl exec busybox nslookup kubernetes-dashboard

---

k8s-UI 安装

• 安装 kubernetes-dashboard.yaml

配置文件需要修改注意地方

-  --apiserver-host=http://192.168.1.32:8080

• 安装 kubernetes-dashboardsvc.yaml

• 访问

http://nodeip:port/
http://master:8080/ui/

---

k8s 应用发布

以发布tomcat 测试

• 安装 my-web-rc.yaml

• 安装 my-web-svc.yaml

• 访问

http://nodeip:port/

---

常用命令集

服务 启动/停止/重启

start 启动 stop 停止 restart 重启

• etcd

systemctl start etcd

• master

systemctl start kube-apiserver.service
systemctl start kube-controller-manager.service
systemctl start kube-scheduler.service

• node

systemctl start kubelet.service
systemctl start kubelet-proxy.service

• flanneld

systemctl restart flanneld

• 重载系统配置

systemctl daemon-reload
systemctl restart docker

---

• 创建

kubectl create -f *.yaml
kubectl apply -f *.yaml

---

• 查看

kubectl get namespances
kubectl get deployment --all-namespaces
kubectl get pods --all-namespaces
kubectl get svc --all-namespaces
kubectl get pod -o wide --all-namespaces
kubectl get pod --namespace=“kuber-system” #指定表空间查询

kubectl describe pod --namespace=“kube-system”
kubectl describe svc --namespace=“kube-system”
kubectl -c <容器名>

kubectl logs -f -n kube-system

---

• 删除

kubectl delete -f *.yaml
kubectl delete svc --namespace=kube-system
kubectl delete pods --grace-period=0 --force
kubectl delete --all pods --namespace=kube-system

---

• cmd

kubectl exec busybox nslookup kubernetes-dashboard
kubectl exec busybox nslookup mytomcat
kubectl exec -i -t busybox /bin/bash

---

• docker 镜像操作

docker tag
docker push
docker pull <地址>
docker serach
docker run [查看网上文档]

---

• etcd

etcdctl set
etcdctl mk
etcdctl update
etcdctl rm
etcdctl rmdir

etcdctl get /k8s/network/config
etcdctl ls

---

• 辅助命令

ip link delete docker0

alias kube=‘kubectl’
alias kubegp=‘kubectl get pods’
alias kubegc=‘kubectl get svc’

---

附镜像地址

可以到阿里云镜像库搜索

docker pull registry.cn-shanghai.aliyuncs.com/hyzhou/busybox
docker pull registry.cn-hangzhou.aliyuncs.com/centos-jdk7-tomcat8/tomcat8:0.0.2

docker pull registry.cn-hangzhou.aliyuncs.com/kube_containers/kubedns-amd64:1.9
docker pull registry.cn-hangzhou.aliyuncs.com/kube_containers/kube-dnsmasq-amd64:1.4.1
docker pull registry.cn-hangzhou.aliyuncs.com/kube_containers/exechealthz-amd64:1.2

k8s master 高可用（稍后）

k8s etcd 集群(稍后)