2019-09-23-k8s-1.15.3-update1.16.0

描述背景：

描述背景：

已经搭建完的k8s集群只有两个worker节点，现在增加work节点到集群。原有集群机器：

|  ip           | 自定义域名         |    主机名 |
|  :----:       |     :----:        |   :----:  |
|                |  master.k8s.io    |  k8s-vip  |
|192.168.0.195    |  master01.k8s.io  |  k8s-master-01|
|192.168.0.197   |  master02.k8s.io  |  k8s-master-02| 
|192.168.0.198   |  master03.k8s.io  |  k8s-master-03|
|192.168.0.199    |  node01.k8s.io    |  k8s-node-01|
|192.168.0.202    |  node02.k8s.io    |  k8s-node-02|
|192.168.0.108    |  node03.k8s.io    |  k8s-node-03|
|192.168.0.111    |  node04.k8s.io    |  k8s-node-04|
|192.168.0.115    |  node05.k8s.io    |  k8s-node-05|

 
初始集群参照 https://duiniwukenaihe.github.io/2019/09/18/k8s-expanded/ 由于上周kubernetes1.15.3爆出漏洞，决定将集群升级为1.16.0 。

#升级master节点

##k8s-master-01节点操作
yum list --showduplicates kubeadm --disableexcludes=kubernetes #查看yum源中可支持版本
  
yum install  kubeadm-1.16.0 kubectl-1.16.0 --disableexcludes=kubernetes。我这里懒人方式直接yum update了
kubeadm version
kubeadm upgrade plan --certificate-renewal=false # 如果不加-certificate-renewal=false将重新生成证书
kubeadm upgrade apply v1.16.0
systemctl daemon-reload
systemctl restart kubelet

##发现 master节点一直not ready

参照： https://github.com/coreos/flannel/pull/1181/commits/2be363419f0cf6a497235a457f6511df396685d4

cat <<EOF > /etc/cni/net.d/10-flannel.conflist
{
  "name": "cbr0",
  "cniVersion": "0.2.0",
  "plugins": [
    {
      "type": "flannel",
      "delegate": {
        "hairpinMode": true,
        "isDefaultGateway": true
      }
    },
    {
      "type": "portmap",
      "capabilities": {
        "portMappings": true
      }
    }
  ]
}
EOF

最好还是修改下kube-flannel.yml

kubectl apply -f kube-flannel.yml

##依次升级其他master02 master03节点

yum install  kubeadm-1.16.0 kubectl-1.16.0 --disableexcludes=kubernetes
kubeadm upgrade apply v1.16.0
systemctl daemon-reload
systemctl restart kubelet

##worker nodes 节点升级

#在master01节点依次将worker节点设置为不可调用
kubectl drain $node --ignore-daemonsets
yum install  kubeadm-1.16.0 kubectl-1.16.0 --disableexcludes=kubernetes
systemctl daemon-reload
systemctl restart kubelet
systemctl status kubelet
kubectl uncordon $node

work节点以此升级

##查看集群## kubectl get nodes

注：碰到的问题早前yaml文件apiVersion用的是extensions/v1beta1 ，1,16要修改为apps/v1beta1。具体看https://kubernetes.io/zh/docs/concepts/workloads/controllers/deployment/

2019-09-23-k8s-1.15.3-update1.16.0

描述背景：

Comments