Kubernetes HA 领导者控制平面服务重启答案

【问题标题】：Kubernetes HA leader control plane services restartingKubernetes HA 领导者控制平面服务重启
【发布时间】：2017-09-13 18:39:11
【问题描述】：

在遵循 kubernetes-the-hard-way 并将其移植到本地 Ubuntu 16.04 虚拟机时，启动 HA 控制平面给我带来了问题。 The kube-apiserver and kube-controller-manager services for the elected leader keep failing and restarting but the control planes on the other two non-leader masters come up fine.此行为跟随集群周围的领导者。

EtcD 看起来很开心：

member 7f44a7567a5e995 is healthy: got healthy result from https://10.1.15.117:2379
member 43d2258c438cbf4e is healthy: got healthy result from https://10.1.15.116:2379
member a83b22e9f907f471 is healthy: got healthy result from https://10.1.15.115:2379
cluster is healthy

我已通过以下方式验证了当前领导者：

kubectl get endpoints kube-controller-manager --namespace=kube-system  -o yaml


apiVersion: v1
kind: Endpoints
metadata:
  annotations:
    control-plane.alpha.kubernetes.io/leader: '{"holderIdentity":"df-dev-kube-test02","leaseDurationSeconds":15,"acquireTime":"2017-09-13T17:15:06Z","renewTime":"2017-09-13T18:10:46Z","leaderTransitions":7}'
  creationTimestamp: 2017-09-13T14:59:57Z
  name: kube-controller-manager
  namespace: kube-system
  resourceVersion: "7814"
  selfLink: /api/v1/namespaces/kube-system/endpoints/kube-controller-manager
  uid: 3599a959-9894-11e7-a34f-005056b24128
subsets: []

kube-apiserver 日志： Pastebin apiserver logs

kube-controller-manager 日志： Pastebin controller logs

【问题讨论】：

标签： kubernetes

【解决方案1】：

考虑到日志中的各种“存储”相关错误，并且由于这与领导选举有关，我认为问题可能与 etcd 有关。不过，Etcd 一直报告很好。出于笑容，我从 api-server.service 中删除了 --experimental-encryption-provider-config=/var/lib/kubernetes/encryption-config.yaml \\，现在控制面板很开心。

我不确定是加密本身造成了破坏，还是我的加密自动化破坏了它。

【讨论】：