005_elasticsearch的数据导入导出

一、全部备份和导入

安装：
git clone https://github.com/taskrabbit/elasticsearch-dump.git
cd elasticsearch-dump
npm install elasticdump -g

sudo yum install npm

(1)创建备份路径
mkdir /data/es_data_backup
(2)迁移原机器上的所有索引到目标机器
#把原始索引的mapping结构和数据导出
elasticdump --input=http://10.200.57.118:9200/ --output=/data/es_data_backup/cmdb_dump-mapping.json --all=true --type=mapping
elasticdump --input=http://10.200.57.118:9200/ --output=/data/es_data_backup/cmdb_dump.json --all=true --type=data

#mapping结构和数据导入新的cluster节点
elasticdump --input=/data/es_data_backup/cmdb_dump-mapping.json --output=http://10.200.57.118:9200/ --bulk=true
elasticdump --input=/data/es_data_backup/cmdb_dump.json --output=http://10.200.57.118:9200/ --bulk=true

二、指定库备份和导入

curl -XGET '192.168.11.10:9200/_cat/indices?v&pretty' .  #查看都有哪些索引
health status index pri rep docs.count docs.deleted store.size pri.store.size
green open jyall-test 5 1 18908740 2077368 25gb 12.5gb

# Backup index data to a file:
elasticdump --input=http://10.200.57.118:9200/ele_nginx_clusters --output=/data/es_data_backup/ele_nginx_clusters_mapping.json --type=mapping

elasticdump --input=http://10.200.57.118:9200/ele_nginx_clusters --output=/data/es_data_backup/ele_nginx_clusters.json --type=data
#或者采用gzip的方式,这种方式亲测节省10多倍的空间,导入时gunzip ele_nginx_clusters.json.gz后再进行导入
#Backup and index to a gzip using stdout:
elasticdump --input=http://10.200.57.118:9200/ele_nginx_clusters --output=$ | gzip > /data/es_data_backup/ele_nginx_clusters.json.gz

导入:
elasticdump --input=/data/es_data_backup/ele_nginx_clusters_mapping.json --output=http://10.200.57.118:9200/ --bulk=true
elasticdump --input=/data/es_data_backup/ele_nginx_clusters.json --output=http://10.200.57.118:9200/ --bulk=true

三、导出遇到的报错及问题

(1)报错如下:
Thu, 26 Apr 2018 09:14:49 GMT | Error Emitted => read ECONNRESET
Thu, 26 Apr 2018 09:14:49 GMT | Total Writes: 19800
Thu, 26 Apr 2018 09:14:49 GMT | dump ended with error (get phase) => Error: read ECONNRESET
(2)
<1>
It sounds like your issue is being caused by the elasticdump opening too many sockets to your elasticsearch cluster. You can use the --maxSockets option to limit the number of sockets opened.
elasticdump --input http://192.168.2.222:9200/index1 --output http://192.168.2.222:9200/index2 --type=data --maxSockets=5

Reference:
https://stackoverflow.com/questions/33248267/dump-ended-with-error-set-phase-error-read-econnreset
https://github.com/nodejs/node/issues/10563

Reference:

https://www.zhangluya.com/?p=543

https://github.com/taskrabbit/elasticsearch-dump