k8s无脑系列（一）-最小集群搭建
修改记录
- Master运行服务中，增加kube-proxy服务
-
Virtual Box
-
CentOS 7
-
yum install docker kubernetes flannel
-
虚拟机网络设置
- 新建虚拟网络A（在主机网络管理菜单），定义网段192.168.56.1/24，打开DHCP
- 所有主机创建两块网卡，外网NAT，内网选择虚拟网络vbox1
- 3台虚拟机设置
| 主机名 | IP | 配置 |
|---|---|---|
| Master | 192.168.56.4 | 1CPU，1G |
| node1 | 192.168.56.5 | 1CPU，1G |
| node2 | 192.168.56.6 | 1CPU，1G |
-
三台主机互通
-
master ssh passwdless 访问其它两台主机
-
打开内核转发
添加net.ipv4.ip_forward 到/etc/sysctl.conf
-
修改Docker配置
修改/usr/lib/systemd/system/docker.service
service段增加 ExecStartPost=/sbin/iptables -I FORWARD -s 0.0.0.0/0 -j ACCEPT
修改/etc/sysconfig/docker
OPTIONS="–insecure-registry gcr.io --selinux-enabled=false"
-
禁用防火墙
systemctl stop firewalld
systemctl disable firewalld -
关闭SElinux
修改/etc/selinux/config
修改SELINUX=disabled -
修改Docker配置，增加镜像
{ "registry-mirrors":["http://hub-mirror.c.163.com"] } -
设置时区与时间同步
yum -y install ntp timedatectl set-timezone Asia/Shanghai - 设置
Master 部分增加了kube-proxy，第一版忘记加了
| 主机名 | 运行的服务 |
|---|---|
| Master | docker etcd kube-apiserver flanneld kube-proxy kube-controller-manager kube-scheduler |
| node1 | docker flanneld kubelet kube-proxy |
| node2 | docker flanneld kubelet kube-proxy |
-
设置etcd参数并启动
vim /etc/etcd/config ETCD_LISTEN_CLIENT_URLS="http://0.0.0.0:2379" # 要注意这个参数的IP，这里因为是连接服务，不是监听，所以必须用IP，不能用0.0.0.0 ETCD_ADVERTISE_CLIENT_URLS="http://192.168.56.4:2379" service etcd restart其实ETCD_LISTEN_CLIENT_URLS怎么设置都无所谓，只要正常监听就可以。放到这里，可以设置为http://192.168.56.4:2379
-
设置flannel
vim /etc/sysconfig/flanneld FLANNEL_ETCD_ENDPOINTS="http://192.168.56.4:2379" FLANNEL_ETCD_PREFIX="/coreos.com/network"稍后补充为什么这么设置
-
手动向etcd，设置flannel网段信息
etcdctl set /coreos.com/network/config '{ "Network": "10.254.0.0/16" }'这里采用kubenetes默认值
-
修改kube-apiserver配置文件
vim /etc/kubernetes/apiserver KUBE_ETCD_SERVERS="--etcd-servers=http://192.168.56.4:2379"稍后测试，是否设置为127.0.0.1会影响node1,2的etcd获取。如果仅仅是影响api-server并且etcd与
api-server在同一台服务器，那么127应该也是可以的。 -
修改controller-manager
vim /etc/kubernetes/controller-manager KUBE_CONTROLLER_MANAGER_ARGS="--node-monitor-grace-period=10s --pod-eviction-timeout=10s" -
编辑kubenetes配置
vim /etc/kubenetes/config KUBE_MASTER="--master=http://192.168.56.4:8080" -
设置服务并启动
systemctl enable docker flanneld kube-apiserver kube-controller-manager kube-scheduler systemctl start docker flanneld kube-apiserver kube-controller-manager kube-scheduler
-
修改kubenetes配置
vim /etc/kubernetes/config KUBE_MASTER="--master=http://192.168.56.4:8080" -
修改kubelets配置
vim /etc/kubernetes/kubelet KUBELET_ADDRESS="--address=0.0.0.0" KUBELET_HOSTNAME="--hostname-override=192.168.56.5" KUBELET_API_SERVER="--api-servers=http://192.168.56.4:8080" -
修改flanneld
vim /etc/sysconfig/flanneld FLANNEL_ETCD_ENDPOINTS="http://192.168.56.4:2379" FLANNEL_ETCD_PREFIX="/coreos.com/network" -
设置服务并启动
systemctl enable docker flanneld kubelet kube-proxy systemctl start docker flanneld kubelet kube-proxy
- 验证节点
$kubectl get nodes -o wide NAME STATUS AGE EXTERNAL-IP 192.168.56.5 Ready 1h <none> 192.168.56.6 Ready 1h <none> - 创建RC文件mysql-rc.yaml
apiVersion: v1 kind: ReplicationController metadata: name: mysql-rc spec: replicas: 1 selector: app: mysql-pod template: metadata: labels: app: mysql-pod spec: containers: - name: mysql image: mysql:5.6 ports: - containerPort: 3306 env: - name: MYSQL_ROOT_PASSWORD value: "123" - 创建Service文件mysql-svc.yaml
apiVersion: v1 kind: Service metadata: name: mysql-svc labels: name: mysql-svc spec: type: NodePort ports: - port: 3306 protocol: TCP targetPort: 3306 name: http nodePort: 30006 selector: app: mysql-pod - 运行
kubectl create -f mysql-rc.yaml
kubectl create -f mysql-svc.yaml
- 结果检查
$kubectl get pods -o wide NAME READY STATUS RESTARTS AGE IP NODE mysql-rc-82g2m 1/1 Running 0 1h 172.17.0.2 192.168.56.5 $kubectl get svc -o wide NAME CLUSTER-IP EXTERNAL-IP PORT(S) AGE SELECTOR kubernetes 10.254.0.1 <none> 443/TCP 2d <none> mysql-svc 10.254.255.244 <nodes> 3306:30006/TCP 1h app=mysql-pod $kubectl get endpoints NAME ENDPOINTS AGE kubernetes 192.168.1.114:6443 2d mysql-svc 172.17.0.2:3306 1h 全部关联上，创建没问题
$ mysql -uroot -p123 -P30006 -h 192.168.56.5
mysql> mysql连接成功