首页 > 代码库 > elasticsearch使用river同步mysql数据
elasticsearch使用river同步mysql数据
====== mysql的river介绍======
- 什么是river?river代表es的一个数据源,也是其它存储方式(如:数据库)同步数据到es的一个方法。它是以插件方式存在的一个es服务,通过读取river中的数据并把它索引到es中,官方的river有couchDB的,RabbitMQ的,Twitter的,Wikipedia的。这里主要研究针对mysql的river。
- mysql的river插件:mysql的river安装见https://github.com/jprante/elasticsearch-river-jdbc,就不具体介绍了。
- 实践测试:
* 环境:
服务器172.16.170.21 数据库:profile 表user
* 创建索引
curl -XPUT ‘http://localhost:9200/profile‘
* 创建数据表与索引映射
curl -XPUT ‘http://localhost:9200/profile/user/_mapping‘ -d ‘
{
"user": {
"properties": {
"id": {
"type": "string",
"store": "yes"
},
"name": {
"type": "string",
"store": "yes"
},
"login_name": {
"type": "string",
"store": "yes"
}
}
}
}‘
* 运行river同步数据
curl -XPUT ‘http://localhost:9200/_river/who_jdbc_river/_meta‘ -d ‘{
"type": "jdbc",
"jdbc": {
"driver": "com.mysql.jdbc.Driver",
"url": "jdbc:mysql://localhost:3306/profile",
"user": "root",
"password": "root",
"sql": "select id as _id,name,login_name from user",
"index": "profile",
"type": "user",
"bulk_size": 100,
"max_bulk_requests": 30,
"bulk_timeout": "10s",
"flush_interval": "5s",
"schedule": "0 0-59 0-23 ? * *"
}
}‘
* 增量更新索引
增量更新,表需要维护时间戳,发现时间戳更新的列需要更新
curl -XPUT ‘http://localhost:9200/_river/who_jdbc_river/_meta‘ -d ‘{
"type": "jdbc",
"jdbc": {
"driver": "com.mysql.jdbc.Driver",
"url": "jdbc:mysql://localhost:3306/profile",
"user": "root",
"password": "root",
"sql": [
{
"statement": "select id as _id,name,login_name from user where mytimestamp > ?",
"parameter": [
"$river.state.last_active_begin"
]
}
],
"index": "profile",
"type": "user",
"bulk_size": 100,
"max_bulk_requests": 30,
"bulk_timeout": "10s",
"flush_interval": "5s",
"schedule": "0 0-59 0-23 ? * *"
}
}‘
删除river
curl -XDELETE ‘localhost:9200/_river/report_jdbc_river‘
elasticsearch使用river同步mysql数据