作者介绍：简历上没有一个精通的运维工程师。请点击上方的蓝色《运维小路》关注我，下面的思维导图也是预计更新的内容和当前进度(不定时更新)。
中间件，我给它的定义就是为了实现某系业务功能依赖的软件，包括如下部分:
Web服务器
代理服务器
ZooKeeper
Kafka
RabbitMQ
Hadoop HDFS
Elasticsearch ES (本章节）
本文章源自我2019年写过的一篇文章：es备份，经过阿里公有云和腾讯私有云的验证，都是可用的，但是经过多年，部分信息已经发生改变，所以基于这个步骤重做步骤，已经验证成功。
Elasticsearch（ES）本身是存储大数据的，一般的情况是没有备份需求的，但是ES还是给我们提供了备份接口和方法。
这里的备份是可以备份到本地，也可以备份到对象存储里面，我这里直接以使用对象存储来作为存储后端。我在历史文章里面使用的是单机的备份，集群的备份也差不多（其实主要我也忘记了）。
这里的备份都是需要插件支持，如果你本地集群则需要自己安装，如果你用的云ES则一般都会自带。
1.插件安装
所有节点都需要安装，我还是以历史记录三节点集群作为演示。
bash
[es@localhost elasticsearch-7.17.29]$ ./bin/elasticsearch-plugin install repository-s3
-> Installing repository-s3
-> Downloading repository-s3 from elastic
[=================================================] 100%
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@ WARNING: plugin requires additional permissions @
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
* java.lang.RuntimePermission accessDeclaredMembers
* java.lang.RuntimePermission getClassLoader
* java.lang.reflect.ReflectPermission suppressAccessChecks
* java.net.SocketPermission * connect,resolve
* java.util.PropertyPermission es.allow_insecure_settings read,write
See https://docs.oracle.com/javase/8/docs/technotes/guides/security/permissions.html
for descriptions of what these permissions allow and the associated risks.
Continue with installation? [y/N]y
-> Installed repository-s3
-> Please restart Elasticsearch to activate any plugins installed
[es@localhost elasticsearch-7.17.29]$ ll plugins/
total 4
drwxr-xr-x 2 es es 4096 Sep 1 23:53 repository-s3
[es@localhost elasticsearch-7.17.29]$ 2.修改配置文件
bash
#这里需要填写oss的ak和sk等信息，修改完成以后启动es
s3.client.default.access_key: xxxxx
s3.client.default.secret_key: xxxx
s3.client.default.endpoint: https://oss-cn-chengdu.aliyuncs.com
s3.client.default.protocol: https
s3.client.default.path_style_access: true3.配置ak和sk
bash
#这个步骤可能不需要，因为我调试的 先做的这个步骤，后做的上面的第二步
#这里需要对应的ak和sk
bin/elasticsearch-keystore add s3.client.default.access_key
bin/elasticsearch-keystore add s3.client.default.secret_key4.创建备份仓库
这里只是创建仓库，并不会真正备份，只需要执行一次即可。es_backup是仓库的名字，可以根据自己需要自己定义。
bash
[es@localhost elasticsearch-7.17.29]$ curl -X PUT 'http://192.168.31.172:9200/_snapshot/es_backup?verify=true' \
> -H 'Content-Type: application/json' \
> -d '{
> "type": "s3",
> "settings": {
> "bucket": "es-bucket123",
> "endpoint": "oss-cn-chengdu.aliyuncs.com",
> "protocol": "https",
> "path_style_access": false,
> "server_side_encryption": false,
> "compress": true
> }
> }'
{"acknowledged":true}[es@localhost elasticsearch-7.17.29]$ 5.查看仓库
bash
[es@localhost elasticsearch-7.17.29]$ curl -X GET 'http://192.168.31.172:9200/_snapshot/es_backup?pretty'
{
"es_backup" : {
"type" : "s3",
"settings" : {
"bucket" : "es-bucket123",
"path_style_access" : "false",
"endpoint" : "oss-cn-chengdu.aliyuncs.com",
"protocol" : "https",
"server_side_encryption" : "false",
"compress" : "true"
}
}
}
[es@localhost elasticsearch-7.17.29]$ 6.执行备份
es_backup：前面创建的快照仓库名称。
snapshot_20250902：快照名称（建议用日期命名）。
wait_for_completion=true：同步等待，直到备份完成才返回（适合小数据量）如果不加 wait_for_completion，则是异步执行
bash
[es@localhost elasticsearch-7.17.29]$ curl -X PUT "http://192.168.31.172:9200/_snapshot/es_backup/snapshot_20250902?wait_for_completion=true"
{"snapshot":{"snapshot":"snapshot_20250902","uuid":"IcIiVPtmSnW6gYnGohfihw","repository":"es_backup","version_id":7172999,"version":"7.17.29","indices":["minute-metrics",".ds-.logs-deprecation.elasticsearch-default-2025.08.01-000001","raw_data-2025.08.16","minute_avg","raw-metrics-2025.08.15-01",".ds-ilm-history-5-2025.08.01-000001",".geoip_databases",".ds-.logs-deprecation.elasticsearch-default-2025.09.01-000002","minute-metrics-2025.08.15",".ds-ilm-history-5-2025.09.01-000002",".kibana-event-log-7.17.29-000001","hour-metrics-2025.08.15","hour-metrics"],"data_streams":["ilm-history-5",".logs-deprecation.elasticsearch-default"],"include_global_state":true,"state":"SUCCESS","start_time":"2025-09-01T17:53:28.868Z","start_time_in_millis":1756749208868,"end_time":"2025-09-01T17:53:39.684Z","end_time_in_millis":1756749219684,"duration_in_millis":10816,"failures":[],"shards":{"total":13,"failed":0,"successful":13},"feature_states":[{"feature_name":"geoip","indices":[".geoip_databases"]}]}}[es@localhost elasticsearch-7.17.29]$ 上面是部分全部索引，我们也可以只备份指定索引，如下面。
bash
curl -X PUT "http://192.168.31.172:9200/_snapshot/es_backup/snapshot_20250902?wait_for_completion=true" \
-H 'Content-Type: application/json' \
-d '{
"indices": "index-2025*,log-nginx",
"ignore_unavailable": true,
"include_global_state": false
}'
7.查看备份结果
bash
[es@localhost elasticsearch-7.17.29]$ curl -X GET "http://192.168.31.172:9200/_snapshot/es_backup/_all?pretty"
{
"snapshots" : [
{
"snapshot" : "snapshot_20250902",
"uuid" : "IcIiVPtmSnW6gYnGohfihw",
"repository" : "es_backup",
"version_id" : 7172999,
"version" : "7.17.29",
"indices" : [
"minute-metrics",
".ds-.logs-deprecation.elasticsearch-default-2025.08.01-000001",
"raw_data-2025.08.16",
"minute_avg",
"raw-metrics-2025.08.15-01",
".ds-ilm-history-5-2025.08.01-000001",
".geoip_databases",
".ds-.logs-deprecation.elasticsearch-default-2025.09.01-000002",
"minute-metrics-2025.08.15",
".ds-ilm-history-5-2025.09.01-000002",
".kibana-event-log-7.17.29-000001",
"hour-metrics-2025.08.15",
"hour-metrics"
],
"data_streams" : [
"ilm-history-5",
".logs-deprecation.elasticsearch-default"
],
"include_global_state" : true,
"state" : "SUCCESS",
"start_time" : "2025-09-01T17:53:28.868Z",
"start_time_in_millis" : 1756749208868,
"end_time" : "2025-09-01T17:53:39.684Z",
"end_time_in_millis" : 1756749219684,
"duration_in_millis" : 10816,
"failures" : [ ],
"shards" : {
"total" : 13,
"failed" : 0,
"successful" : 13
},
"feature_states" : [
{
"feature_name" : "geoip",
"indices" : [
".geoip_databases"
]
}
]
}
],
"total" : 1,
"remaining" : 0
}
[es@localhost elasticsearch-7.17.29]$ 8.查看oss里面的数据
9.还原
除了备份哪个操作换成下面的还原命令，其他都是一样的。
curl -XPOST 192.168.2.23:9200/_snapshot/es_backup/es_bak1/_restore?wait_for_completion=true运维小路
一个不会开发的运维！一个要学开发的运维！一个学不会开发的运维！欢迎大家骚扰的运维！
关注微信公众号《运维小路》获取更多内容。