ElasticSearch数据导入

作者&＃xff1a;dominating

本文将介绍通过logstash收集.csv文件&＃xff0c;oracle数据库数据再导入到ElasticSearch中&＃xff0c;以及SuperMap iClient for Javascript 9D与ElasticSearch的结合使用。
###安装logstash
1、官网下载logstash-6.1.2.tar.gz&＃xff0c;https://www.elastic.co/downloads/logstash
解压到/opt下
2、测试
运行logstash

bin/logstash -e &＃39;input { stdin { } } output { stdout {} }&＃39;


输入hello world&＃xff0c;logstash将会输出内容到控制台

2018-01-29T16 :36:49,507 &＃43;0000 0.0.0.0 hello world


###导入.csv文件
编写配置文件
收集和导入数据需要借助logstash的input,filter,output插件来编写配置文件:

input {file {path &＃61;> ["/opt/flights2.csv"]start_position &＃61;> "beginning"}
}
filter {csv {separator &＃61;> ","columns &＃61;> ["ident","lon","lat","temp","origin","destination"]}
}
output {elasticsearch {hosts &＃61;> ["192.168.255.143:9200"]index &＃61;> "flight"}
}

参数说明&＃xff1a;
input插件
file&＃xff1a;数据来源为文件型
path:#必选项&＃xff0c;配置文件路径.如我使用的以下.csv文件

start_position&＃xff1a;logstash从哪个位置读取文件数据&＃xff0c;默认从尾部&＃xff0c;值为&＃xff1a;end&＃xff0c;如果要导入历史数据则设置成&＃xff1a;beginning
filter插件
csv:csv文件过滤器
separator&＃xff1a;定义列分割符值。默认为逗号’,’
columns&＃xff1a;定义一个列名称列表&＃xff0c;按照在CSV中出现的顺序
output插件
elasticsearch输出目标为elasticsearch&＃xff0c;配置host和index索引名

运行logstash

/opt/logstash-6.1.2/bin/logstash -f /opt/test.conf


-f&＃xff1a;指定配置文件路径

查看ElasticSearch

"took": 9,"timed_out": false,"_shards": {"total": 5,"successful": 5,"skipped": 0,"failed": 0},"hits": {"total": 100,"max_score": 1.0,"hits": [{"_index": "flight","_type": "doc","_id": "9pMPQWEBy48LTNYe0eDu","_score": 1.0,"_source": {"temp": "1.49E&＃43;12","&＃64;timestamp": "2018-01-29T08:36:50.287Z","host": "ubuntu-node3","lat": "29.824944","ident": "T0000","origin": "Lishe","destination": "Jiangbei","message": "T0000,121.465069,29.824944,1.49E&＃43;12,Lishe,Jiangbei\r","&＃64;version": "1","path": "/opt/flights2.csv","lon": "121.465069"}}]}

可以看到导入了100条数据&＃xff0c;并且能够被查询到。
###导入oracle数据
需要在logstash所在机器上提前安装oracle客户端&＃xff0c;并且配置好oracle的环境变量
测试oracle数据库是否能正常通讯

编写logstash配置文件
需要导入的oracle数据

input {jdbc {jdbc_driver_library &＃61;> "/opt/ojdbc6.jar"jdbc_driver_class &＃61;> "Java::oracle.jdbc.driver.OracleDriver"jdbc_connection_string &＃61;> "jdbc:oracle:thin:&＃64;//192.168.15.89:1521/supermap"jdbc_user &＃61;> "liu"jdbc_password &＃61;> "supermap"schedule &＃61;> "* * * * *"statement &＃61;> "select * from SMDTV_341"type &＃61;> "jdbc"last_run_metadata_path &＃61;> "/home/elsearch/logstash-oradb.lastrun"}
}
filter {}
output {elasticsearch {hosts &＃61;> ["192.168.255.143:9200"]index &＃61;> "test"}
}


参数说明&＃xff1a;
jdbc_driver_library&＃xff1a;在oracle目录下&＃xff0c;如我的在D:\app\wangwu\product\11.2.0\dbhome_1\jdbc\lib下面&＃xff0c;复制到指定目录

schedule&＃xff1a;查询间隔&＃xff0c;"* * * * *"每分钟查询一次&＃xff0c;不设置则只执行一次
last_run_metadata_path&＃xff1a;最后更新时间文件位置
statement&＃xff1a;SQL查询语句
index&＃xff1a; 索引&＃xff0c;可以先创建一个索引再导入

将文件保存为jdbc.conf放在logstash所在机器&＃xff0c;运行

/opt/logstash-6.1.2/bin/logstash -f /opt/jdbc.conf


查看结果

说明数据已经导入成功&＃xff0c;并且能够被查询到。
###导入地理坐标点数据
ElasticSearch提供了地理位置功能&＃xff0c;并且能够把地理位置、全文搜索、结构化搜索和分析结合到一起。
ElasticSearch中存储地理坐标数据需要使用geo-point类型&＃xff0c;并且必须提前显式声明
1、创建索引&＃xff0c;并且声明location为geo-point类型

curl -XPUT &＃39;192.168.255.143:9200/test3?pretty&＃39; -H &＃39;Content-Type: application/json&＃39; -d&＃39;
{ "mappings": { "capital" : { "properties" : { "location" : { "type" : "geo_point" } } } }
}
&＃39;

创建索引test3&＃xff0c;映射字段location&＃xff0c;并且声明为geo-point类型
2、编写配置文件
将带地理坐标的数据导入到ElasticSearch中

input { jdbc { jdbc_driver_library &＃61;> "/opt/ojdbc6.jar" jdbc_driver_class &＃61;> "Java::oracle.jdbc.driver.OracleDriver" jdbc_connection_string &＃61;> "jdbc:oracle:thin:&＃64;//192.168.15.89:1521/supermap" jdbc_user &＃61;> "liu" jdbc_password &＃61;> "supermap" statement &＃61;> "select * from SMDTV_361" type &＃61;> "jdbc" last_run_metadata_path &＃61;> "/home/elsearch/logstash-oradb.lastrun" }
}
filter { mutate { add_field &＃61;> {"location" &＃61;> "%{smy},%{smx}"} }
}
output { elasticsearch { hosts &＃61;> ["192.168.255.143:9200"] index &＃61;> "test3" document_type &＃61;> "capital" }
}

由于原表中并没有location字段&＃xff0c;所以filter中使用mutate插件给收集到的数据添加字段location
3、执行导入命令

/opt/logstash-6.1.2/bin/logstash -f /opt/jdbc.conf


4、网格聚合
将地理位置数据导入成功之后&＃xff0c;我们就可以使用ElasticSearch提供的地理位置功能了&＃xff0c;以下我们将执行Geohash网格聚合&＃xff1a;

curl -XGET &＃39;192.168.255.143:9200/test3/capital/_search?pretty&＃39; -H &＃39;Content-Type: application/json&＃39; -d&＃39;
{ "query": { "constant_score": { "filter": { "geo_bounding_box": { "location": { "top_left": { "lat": 90, "lon": -180 }, "bottom_right": { "lat": -90, "lon": 180 } } } } } }, "aggs": { "world": { "geohash_grid": { "field": "location", "precision": 1 } } }
}
&＃39;

如果对Geohash不太明白&＃xff0c;可以参考官方的文档&＃xff1a;https://www.elastic.co/guide/cn/elasticsearch/guide/current/geohash-grid-agg.html
###SuperMap iClient for Javascript 9D和ElasticSearch的结合使用
SuperMap iClient for Javascript 9D封装了ElasticSearch的Javascript API&＃xff0c;我们以for Leaflet为例查询之前导入的数据。
1、 定义服务

liveESService &＃61; new SuperMap.ElasticSearch("http://192.168.255.143:9200");


2、 传入查询条件&＃xff0c;成功回调函数

function loadLiveData() {var liveParameters &＃61; [];liveParameters.push({index: "flight"});liveParameters.push({"query": {"match_all":{}},"from": 0,"size": 100});liveESService.msearch({body: liveParameters}, function (error, result) {if (error) {widgets.alert.showAlert(JSON.stringify(error), false);return;}renderLive(result.responses);});}

使用new SuperMap.ElasticSearch(url).msearch(params,callback)方法&＃xff0c;传入查询参数&＃xff0c;定义回调函数
3、 处理数据添加到地图

4、更多关于SuperMap iClient for Javascript 9D和ElasticSearch的结合使用的例子
http://iclient.supermap.io/examples/leaflet/examples.html#Elasticsearch
热力/网格图

航班监控