1.在master或者node节点 停止服务service kube-calico stop
2.然后重启失败 service kube-calico start
3.查看报错：
journalctl -f -u kube-calico
报错：container name "/calico-node" is already in use by container "43fa89620e14
4.查看日志：
root@blk:/home/k8s# docker logs 0152f90297
2019-04-18 06:38:05.992708 I | warning: ignoring ServerName for user-provided CA for backwards compatibility is deprecated
Skipping datastore connection test
Using IPv4 address from environment: IP=172.16.5.201
IPv4 address 172.16.5.201 discovered on interface ens3
Calico node 'k8snode3' is already using the IPv4 address 172.16.5.201:
- Check the node configuration to remove the IP address conflict
Terminating
Calico node failed to start
提示k8snode3已经存在
5.root@blk:/home/k8s# calicoctl get node
NAME
k8snode3
peer1.org1.example.com
6.删除存在的名字
root@blk:/home/k8s# calicoctl delete nodes k8snode3
Successfully deleted 1 'node' resource(s)
root@blk:/home/k8s# calicoctl delete nodes peer1.org1.example.com
Successfully deleted 1 'node' resource(s)
7.然后删除对应的容器
root@blk:/home/k8s# docker rm -f 43fa89620e14
43fa89620e14
8.最后重启服务
root@blk:/home/k8s# service kube-calico start
root@blk:/home/k8s# service kube-calico status
● kube-calico.service - calico node
Loaded: loaded (/lib/systemd/system/kube-calico.service; enabled; vendor preset: enabled)
Active: active (running) since Thu 2019-04-18 06:42:12 UTC; 2s ago
Process: 26367 ExecStop=/usr/bin/docker rm -f calico-node (code=exited, status=0/SUCCESS)
Main PID: 31858 (docker)
Tasks: 9 (limit: 4915)
CGroup: /system.slice/kube-calico.service
kubectl get nodes：node 列表
kubectl delete node xxx.xx.xxx-ip：删除 inactive 的 dead node （通过 ip）
calicoctl get nodes：calico node 列表
calicoctl delete node xxx: 删除 dead node （通过 hostname）
calicoctl node run: 在新节点上跑 calico node ，让新的 node 生效 （bgp peer）
calicoctl node status: 查看状态，确认没有问题
网友评论