K8S HA高可用切换测试问题

lentil1016 / kubeadm-ha

Deprecated! Boot a ha kubernetes 1.11.0/1.12.1/1.13.0/1.14.0 cluster with kubeadm.

GNU General Public License v3.0

214 stars 128 forks source link

K8S HA高可用切换测试问题 #28

Open guoh1988 opened 5 years ago

guoh1988 commented 5 years ago

按照文档部署集群 1.使用power off 和shutdown关闭管理节点apiserver会在10S内切换。 2.物理机直接关闭电源虚拟机直接通过控制台关闭电源，会出现apiserver5min以上不可用，虚拟IP可以正常切换（关闭kube-controller-manager，kube-scheduler所选主的那个节点）还请帮忙确认下

lentil1016 commented 5 years ago

I'll look into it ASAP.

lentil1016 commented 5 years ago

可以的话请提供以下信息：

你是如何判断ApiServer不可用的，执行什么命令时出了错误，日志里何处出现异常
使用的是哪个版本的脚本
执行kubectl get cs的结果

guoh1988 commented 5 years ago

使用的是1.11.0的版本在剩下的两个master节点上直接使用kubectl get nodes/cs/ep/pods 查看基本信息会出现查询超时

BigDataHunter commented 5 years ago

error execution phase preflight: [preflight] Some fatal errors occurred: [ERROR Port-10251]: Port 10251 is in use [preflight] If you know what you are doing, you can make a check non-fatal with --ignore-preflight-errors=... cp: cannot stat ‘/etc/kubernetes/admin.conf’: No such file or directory unable to recognize "STDIN": Get https://192.168.207.155:8443/api?timeout=32s: dial tcp 192.168.207.155:8443: connect: connection refused

jacinchan commented 5 years ago

i use similar solution. it seems keepalived LVS cause this problem config keepalived as "all backup, nopreempt" can fix this