安装2个 master ,2 个node的集群，etcd安装在master1和两个node上，当master1宕机后，集群不可用，而 master2宕机后，集群可用；

[fmonkeyaksd]

What happened? 发生了什么问题？

master1宕机后， systemctl status kube-apiserver报错： addConn.createTransport failed to connect to {https://10.4.29.213:2379}

What did you expect to happen? 期望的结果是什么？

两个master任意一个宕机，集群均正常。

How can we reproduce it (as minimally and precisely as possible)? 尽可能最小化、精确地描述如何复现问题

2个 master ,2 个node的集群，etcd安装在master1和两个node上，当master1宕机后，集群不可用，而 master宕机后，集群可用

Anything else we need to know? 其他需要说明的情况

No response

Kubernetes version k8s 版本

1.17.2

Kubeasz version

2.0.0

OS version 操作系统版本

```console # On Linux: $ cat /etc/os-release NAME="Red Hat Enterprise Linux Server" VERSION="7.3(Ma·po）" ID="rhel" ID LIKE="fedora" VERSION ID="7.3" PRETTY NAME="RedHat Enterprise Linux Server 7.3 (Maipo)" ANSI COLOR="0;31" CPE_NAME="cpe:/o:redhat:enterprise_linux:7.3:GA:server"HOME URL="https://www.redhat.com/ BUG REPORT URL="https://bugzilla.redhat.com/" REDHAT BUGZILLA PRODUCT="Red Hat Enterprise Linux 7REDHAT BUGZILLA PRODUCT_VERSION=7.3 REDHAT_SUPPORT_PRODUCT="Red Hat Enterprise Linux" REDHAT SUPPORT PRODUCT_VERSION=7 $ uname -a Linux 12922133.10.0-514.el7.x86_64 #1 SMP Wed Oct 19 11:24:13 EDT 2016 x86_64 x86_64 x86_64 GNU/Linux

Related plugins (CNI, CSI, ...) and versions (if applicable) 其他网络插件等需要说明的情况

[gjmzj]

部署节点是master1吗？你可以试试把部署节点独立出来；因为master1宕机，是不是部署节点没有了，以为集群不可用了

[fmonkeyaksd]

我定位问题原因是，在集群安装过程中，计算节点的/etc/kubelet.kubeconfig和/etc/kube-proxy.kubeconfig里配置的apiserver是集群默认的第一个master节点:6443端口，这就导致当第一个master挂掉后，计算节点无法连接apiserver

[im-jinxinwang]

@fmonkeyaksd 两个etcd？ 两个etcd没有高可用性。

[github-actions[bot]]

This issue is stale because it has been open for 30 days with no activity.

[github-actions[bot]]

This issue was closed because it has been inactive for 14 days since being marked as stale.