ETCD单点故障恢复

问题描述

etcd挂载目录硬盘损坏，/var/lib/etcd目录丢失，导致master1上的etcd异常
硬盘恢复后，重新创建逻辑卷，创建目录，重新挂载

解决处理

挂载后恢复etcd

# 1. 首先停止etcd服务，否则会一直生成data目录
systemctl stop etcd
# 2. 移除data目录，新建data目录
cd /var/lib/etcd
mv data data.bak
mkdir data
# 3. 修改etcd服务配置文件
vim /etc/systemd/system/etcd.service
--initial-cluster-state=existing \
# 4. 重启etcd
systemctl daemon-reload
systemctl restart etcd
# 5. 检查etcd集群状态
ETCDCTL_API=3 etcdctl endpoint status -w table --endpoints ip屏蔽.ip屏蔽.ip屏蔽.1:2379,ip屏蔽.ip屏蔽.ip屏蔽.2:2379,ip屏蔽.ip屏蔽.ip屏蔽.3:2379 --cert="/var/lib/etcd/certs/etcd.pem"  --key="/var/lib/etcd/certs/etcd-key.pem" --cacert="/var/lib/etcd/certs/ca.pem"

---

etcdctl文件见【ETCD备份恢复文章】↓

ETCD备份恢复

 ## 备份 1. 创建测试pod ```bash kubectl run test --image=nginx:1.7.9 --labels="app=test" kubectl run test --image=nginx:1.7.9 --labels="app=test" -n kube-system ``` 2. 在