在centos7上安装配置k8s集群

2017-09-13 10:31:41来源:http://blog.yubangweb.com/k8sshi-jian-bi-ji/作者:yubang的博客人点击

分享
配置背景介绍

为什么要用kubernetes这么复杂的docker集群管理工具呢?一开始接触了docker内置的swarm,这个工具非常简单快捷的完成docker集群功能。但是在使用docker1.13内置的swarm做集群的时候遇到vip负载均衡没有正确映射端口到外网,或者出现地址被占用的情况,这对高可用性的需求是不利的,然而又没找到一个解决方案,只能转投k8s。


实验环境
腾讯云
centos7.3 64位
安装
yum-config-manager --add-repo https://docs.docker.com/v1.13/engine/installation/linux/repo_files/centos/docker.repo
yum makecache fast
yum -y install docker-engine-1.13.1
yum install epel-release -y
yum remove -y docker-engine*
yum install -y kubernetes etcd docker flannel
修改配置文件

注意下面的10.135.163.237换成自己服务器ip


sed -i "s/localhost:2379/10.135.163.237:2379/g" /etc/etcd/etcd.conf
sed -i "s/localhost:2380/10.135.163.237:2380/g" /etc/etcd/etcd.conf
sed -i "s/10.135.163.237:2379/10.135.163.237:2379,http:////127.0.0.1:2379/g" /etc/etcd/etcd.conf
sed -i "s/127.0.0.1:2379/10.135.163.237:2379/g" /etc/kubernetes/apiserver
sed -i "s/--insecure-bind-address=127.0.0.1/--insecure-bind-address=0.0.0.0/g" /etc/kubernetes/apiserver
sed -i "s/--admission-control=NamespaceLifecycle,NamespaceExists,LimitRanger,SecurityContextDeny,ServiceAccount,ResourceQuota/--admission_control=NamespaceLifecycle,NamespaceExists,LimitRanger,ResourceQuota/g" /etc/kubernetes/apiserver
sed -i "s/--hostname-override=127.0.0.1/--hostname-override=10.135.163.237/g" /etc/kubernetes/kubelet
sed -i "s/127.0.0.1:8080/10.135.163.237:8080/g" /etc/kubernetes/kubelet
sed -i "s/--address=127.0.0.1/--address=0.0.0.0/g" /etc/kubernetes/kubelet
sed -i "s/127.0.0.1:8080/10.135.163.237:8080/g" /etc/kubernetes/config
sed -i "s/127.0.0.1:2379/10.135.163.237:2379/g" /etc/sysconfig/flanneld
修改host
vi /etc/hosts
10.135.163.237 k8s_master
添加网络
systemctl enable etcd.service
systemctl start etcd.service
etcdctl mk //atomic.io/network/config '{"Network":"172.17.0.0/16"}'
启动服务
service docker start
for SERVICES in etcd kube-apiserver kube-controller-manager kube-scheduler kube-proxykubelet docker flanneld ; do systemctl restart $SERVICES; systemctl enable $SERVICES; systemctl status $SERVICES; done;
第一个demo

编写文件a.yaml


apiVersion: extensions/v1beta1
kind: Deployment
metadata:
name: my-app
spec:
replicas: 2
template:
metadata:
labels:
app: my-app
spec:
containers:
- name: my-app
image: registry.alauda.cn/yubang/paas_base_test
ports:
- containerPort: 80
command: ["/bin/bash", "/var/start.sh"]
resources:
limits:
cpu: 0.5
memory: 64Mi

编写文件b.yaml


apiVersion: v1
kind: Service
metadata:
name: my-app-svc
labels:
app: my-app
spec:
ports:
- port: 80
targetPort: 80
nodePort: 30964
type: NodePort
selector:
app: my-app

创建服务


kubectl create -f a.yaml --validate
kubectl create -f b.yaml --validate

删除服务


kubectl delete -f a.yaml
kubectl delete -f b.yaml
增加子节点

安装软件


yum-config-manager --add-repo https://docs.docker.com/v1.13/engine/installation/linux/repo_files/centos/docker.repo
yum makecache fast
yum -y install docker-engine-1.13.1
yum install epel-release -y
yum remove -y docker-engine*
yum install -y kubernetes docker flannel

修改配置文件(10.135.163.237为主节点ip,139.199.0.29为当前节点ip)


sed -i "s/--hostname-override=127.0.0.1/--hostname-override=139.199.0.29/g" /etc/kubernetes/kubelet
sed -i "s/127.0.0.1:8080/10.135.163.237:8080/g" /etc/kubernetes/kubelet
sed -i "s/--address=127.0.0.1/--address=0.0.0.0/g" /etc/kubernetes/kubelet
sed -i "s/127.0.0.1:8080/10.135.163.237:8080/g" /etc/kubernetes/config
sed -i "s/127.0.0.1:2379/10.135.163.237:2379/g" /etc/sysconfig/flanneld
sed -i "s/--admission-control=NamespaceLifecycle,NamespaceExists,LimitRanger,SecurityContextDeny,ServiceAccount,ResourceQuota/--admission_control=NamespaceLifecycle,NamespaceExists,LimitRanger,ResourceQuota/g" /etc/kubernetes/apiserver

启动服务


service docker start
for SERVICES in kube-proxy kubelet docker flanneld; do
systemctl restart $SERVICES
systemctl enable $SERVICES
systemctl status $SERVICES
done;

在主服务器查看节点


kubectl get node

重启服务(重新加入集群)


systemctl restart kube-apiserver.service

删除节点


kubectl delete node 节点ip

最新文章

123

最新摄影

微信扫一扫

第七城市微信公众平台