etcd :备份与恢复

背景：
对于k8s集群，etcd集群的健康非常关键，我们要对etcd进行备份，以免出现问题
本文，代码，针对容器化部署的k8s (kubeadm安装方式）
备份：
脚本名： etcd-backup.sh

#!/bin/bash
#backup etcd data
set -e

etcdurl=`docker exec etcd ps -o args |grep -v ps |grep -v COMMAND |awk '{print $7}'`
health=`docker exec etcd etcdctl --endpoints ${etcdurl} cluster-health |grep 'cluster is healthy'`
if [ "$health" = "cluster is healthy" ]; then
  echo cluster is healthy
  backuptime=`date +%F-%H-%M-%S`
  etcdbackup=`docker exec -e ETCDCTL_API=3 etcd etcdctl --endpoints ${etcdurl} snapshot save snapshotdb-${backuptime} |awk '{print $4}'`
  mv ${etcdbackup}   /data/backup/etcd-backup
  echo cluster backup is done!
else
  echo cluster is not healthy
  exit 1
fi

测试
运行之后就可以在备份存储路径下面看到 新的备份文件
注意：代码默认的备份路径为：/data/backup/etcd-backup
可自行修改

编写定时任务
crontab -e 编辑

恢复：
脚本
#注意手动修改最近备份文件名为snapshot.db

#!/bin/bash
#restore etcd data
set -e

etcdurl=`docker exec etcd ps -o args |grep -v ps |grep -v COMMAND |awk '{print $7}'`
health=`docker exec etcd etcdctl --endpoints ${etcdurl} cluster-health |grep 'cluster is healthy'`
if [ "$health" = "cluster is healthy" ]; then
  echo "cluster is healthy"
  #注意手动修改最近备份文件名为snapshot.db
  docker exec etcd etcdctl  --endpoints ${etcdurl} snapshot restore snapshot.db   --data-dir=/var/lib/etcd
  echo "etcd restore is done!"
else
  echo "cluster is not healthy"
  exit 1
fi