Kubernetes(k8s)安装

环境准备

# 关闭CentOS防火墙
systemctl disable firewalld
systemctl stop firewalld
# 安装etcd和kubernetes软件
yum install -y etcd kubernetes
# 启动服务
systemctl start etcd
systemctl start docker

如果docker启动失败，请参考(v1 /etc/sysconfig/selinux 把selinux后面的改为disabled,重启一波机器，再重启docker就可以了)

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

配置

Tomcat

mkdir /usr/local/k8s
cd /usr/local/k8s/
vi mytomcat-rc.ymal

mytomcat.rc.yaml

apiVersion: v1
kind: ReplicationController
metadata:
 name: mytomcat
spec: 
 replicas: 2
 selector:
  app: mytomcat
template:
 metadata:
  labels:
  app: mytomcat
 spec:
  containers:
   - name: mytomcat
     image: tomcat:7-jre7
     ports:
     - containerPort: 8088

kubecl create -f mytomcat.rc.yaml

mytomcat.svc.yaml

apiVersion: v1
kind: Service
metadata:
 name: mytomcat
spec:
 type: NodePort
 ports:
  -port: 8888
   nodePort: 30001
 selector:
  app: mytomcat

kubectl create -f mytomcat.svc.yaml

问题解决
docker pull失败
解决方案1

1. yum install rhsm -y
2. docker pull registry.access.redhat.com/thel7/pod-infrastructure:latest
   如果以上两步解决问题了，那么就不需要在执行下面操作
3. docker search pod-infrastructure
4. docker pull docker.io/tianyebj/pod-infrastructure
5. docker tag tianyebj/pod-infrastructure 192.168.126.143:5000/pod-infrastructure
6. docker push 192.168.126.143:5000/pod-infrastructure
7. vi /etc/kubernetes/kubelet
   修改KUBELET_ POD INFRA_CONTAINE="–pod-infra-container-image=192.168.126.143:5000/pod- infrastructure:latest"
8. 重启服务
   systemctl restart kube-apiserver systemctl restart kube-controller-manager systemctl restart kube-scheduler systemctl restart kubelet systemctl restart kube-proxy

解决方案2

1. docker pull kubernetes/pause
2. docker tag docker.io/kubermetes/pause:latest 192.168.126. 143:5000/google_ containers/pause-amd64.3.0
3. docker push 192.168.126.143:5000/google_ containers/pause-amd64.3.0
4. vi /etc/kubernetes/kubelet配置为
   KUBELET_ ARGS="–pod_ jinfra_ container_image=192.168.126.143:5000/google_containers/pause-amd64.3.0"
5. 重启kubelet服务systemctl restart kubelet

外部网不能访问
在搭建好的k8s集群内创建的容器，只能在其所在的节点上curI可访问，但是在其他任何主机上无法访问
容器占用的端口
解决方案: .

1. vim /etc/sysctl.conf
2. net.ipv4.ip. _forward=1

解决kubect get pods时No resources found问题

1. vim /etc/kubernetes/apiserver
2. 找到KUBE_ ADMISSION_ CONTROL="-admission_control=NamespaceLifecycle,Namespacexists,LimitRanger,SecurityContextDeny,ServiceAccount,ResourceQuota",去掉ServiceAccount, 保存退出。
3. systemct restart kube-apiserver重启此服务
   

Kubernetes集群

环境准备与规划
推荐配置2核2G
Docker version 1 7.05.0-ce

角色	IP	组件
master	192.168.126.140	etcd、kube-apiserver、kube-controller-manager、 kube-scheduler、docker
node01	192.168.126.141	kube-proxy. kubelet. docker
node02	192.168.126.142	kube-proxy. kubelet. docker

• 查看默认防火墙状态(关闭后显示not running ,开启后显示running)
  firewall-cmd --state
• 关闭防火墙
  systemctl stop firewalld.service
• 禁止firewal开机启动
  systemctl disable firewalld.service
• 获取Kubernetes二进制