当前位置：运维

K8S v1.10 HA高可用集群部署[阿里云镜像源]

实验环境：

  1. 3台centos 1611版本虚拟机，mini安装。Linux localhost 3.10.0-514.el7.x86_64 #1 SMP Tue Nov 22 16:42:41 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux

  2. docker version 1.13.1

  3. etcd Version: 3.1.13

  4. kubeadm，kubelet，kubectl，kube-cni版本如下：

    kubelet-1.10.0-> 0.x86_64.rpm
kubeadm-1.10.0-0.x86_64.rpm
kubectl-1.10.0-> 0.x86_64.rpm
kubernetes-cni-0.6.0-0.x86_64.rpm

  5. k8s网络组件：flannel:v0.10.0-amd64

  6. 实验网络规划如下：

    host1 172.18.0.154/22 
host2 172.18.0.155/22
host3 172.18.0.156/22
VIP 172.18.0.192/22

  7. 视频教程

安装步骤：

  1. 下载脚本

  2. 在3台主机中执行基础环境配置脚本 base-env-config.sh

  3. 在主机1执行脚本 host1-base-env.sh

  4. 在主机2执行脚本 host2-base-env.sh

  5. 在主机3执行脚本 host3-base-env.sh

  6. 在host1主机执行如下命令

    [root@host1~]# ll /etc/etcd/ssl/
total 12
-rw-r--r--. 1 root root 1387 Mar 30 16:04 ca.pem
-rw-------. 1 root root 1675 Mar 30 16:04 etcd-key.pem
-rw-r--r--. 1 root root 1452 Mar 30 16:04 etcd.pem
[root@host1 ~]# scp -r /etc/etcd/ssl root@172.18.0.155:/etc/etcd/
[root@host1 ~]# scp -r /etc/etcd/ssl root@172.18.0.156:/etc/etcd/

  7. 在3台主机中分别执行脚本 etcd.sh

  8. 查看keepalived状态。如下图所示，则三个节点机之间心跳正常。

    systemctl status keepalived

  9. 查看etcd运行状态。在host1,host2,host3分别执行如下命令：

    etcdctl  --endpoints=https://${NODE_IP}:2379  --ca-file=/etc/etcd/ssl/ca.pem  --cert-file=/etc/etcd/ssl/etcd.pem  --key-file=/etc/etcd/ssl/etcd-key.pem cluster-health

  10. 在3台主机上安装kubeadm,kubelet,kubctl,docker

    yum install kubelet kubeadm kubectl kubernetes-cni docker -y

  11. 在3台主机禁用docker启动项SELinux

    sed -i 's/--selinux-enabled/--selinux-enabled=false/g' /etc/sysconfig/docker

  12. 在3台主机的kubelet配置文件中添加如下参数

    sed -i '9a\Environment="KUBELET_EXTRA_ARGS=--pod-infra-container-image=registry.cn-hangzhou.aliyuncs.com/osoulmate/pause-amd64:3.0"' /etc/systemd/system/kubelet.service.d/10-kubeadm.conf

  13. 在3台主机添加docker加速器配置（可选）

    cat < /etc/docker/daemon.json
{
"registry-mirrors": ["https://wcmntott.mirror.aliyuncs.com"] 
}
EOF

  14. 在3台主机分别执行以下命令

    systemctl daemon-reload
systemctl enable docker && systemctl restart docker
systemctl enable kubelet && systemctl restart kubelet

  15. 在3台主机中分别执行kubeadmconfig.sh生成配置文件config.yaml

  16. 在host1主机中首先执行kubeadm初始化操作,命令如下：

    kubeadm init --config config.yaml

  17. 执行初始化后操作

    mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config

  18. 将主机host1中kubeadm初始化后生成的证书拷贝至host2,host3相应目录下

    scp -r /etc/kubernetes/pki root@172.18.0.155:/etc/kubernetes/
scp -r /etc/kubernetes/pki root@172.18.0.156:/etc/kubernetes/

  19. 为主机host1安装网络组件 podnetwork【这里选用flannel】

    kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml
systemctl stop kubelet
systemctl restart docker
docker pull registry.cn-hangzhou.aliyuncs.com/osoulmate/flannel:v0.10.0-amd64
systemctl start kubelet

  20. 在host2,host3上执行kubeadm初始化操作

    kubeadm init --config config.yaml
kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml
systemctl stop kubelet
systemctl restart docker
docker pull registry.cn-hangzhou.aliyuncs.com/osoulmate/flannel:v0.10.0-amd64
systemctl start kubelet

  21. 查看集群各节点状态

    [root@localhost ~]# kubectl get nodes
NAME      STATUS    ROLES     AGE       VERSION
host1     Ready     master    5m        v1.10.0
host2     Ready     master    1m        v1.10.0
host3     Ready     master    1m        v1.10.0
[root@localhost ~]# kubectl get po --all-namespaces
NAMESPACE     NAME                            READY     STATUS    RESTARTS   AGE
kube-system   coredns-7997f8864c-k9dcx        1/1       Running   0          5m
kube-system   coredns-7997f8864c-sv9rv        1/1       Running   0          5m
kube-system   kube-apiserver-host1            1/1       Running   1          4m
kube-system   kube-apiserver-host2            1/1       Running   0          1m
kube-system   kube-apiserver-host3            1/1       Running   0          1m
kube-system   kube-controller-manager-host1   1/1       Running   1          4m
kube-system   kube-controller-manager-host2   1/1       Running   0          1m
kube-system   kube-controller-manager-host3   1/1       Running   0          1m
kube-system   kube-flannel-ds-88tz5           1/1       Running   0          1m
kube-system   kube-flannel-ds-g9dpj           1/1       Running   0          2m
kube-system   kube-flannel-ds-h58tp           1/1       Running   0          1m
kube-system   kube-proxy-6fsvq                1/1       Running   1          5m
kube-system   kube-proxy-g8xnb                1/1       Running   1          1m
kube-system   kube-proxy-gmqv9                1/1       Running   1          1m
kube-system   kube-scheduler-host1            1/1       Running   1          5m
kube-system   kube-scheduler-host2            1/1       Running   1          1m
kube-system   kube-scheduler-host3            1/1       Running   0          1m

  22. 高可用验证,将host1关机，在host3上执行

    while true; do  sleep 1; kubectl get node;date; done

在host2上观察keepalived是否已切换为主状态。

评论
一些有趣的事儿
热爱 2021-11-02 21:22:34
不错