第三步:下载并安装插件 (插件非常多,以下列出我喜欢的,可以有选择性的安装)
(1) marvel
远程安装方式:
bin/plugin -i elasticsearch/marvel/latest
本地安装方式:
wget https://download.elasticsearch.org/elasticsearch/marvel/marvel-latest.zip
bin/plugin -i marvel -u file:/home/elasticsearch-1.5.1/marvel-latest.zip
在启动后,可以通过以下方式查看elasticsearch运行情况
http://xxx.xxx.xxx.xxx:8765/_plugin/marvel/
(2) elasticsearch service [非常喜欢]
https://github.com/elastic/elasticsearch-servicewrapper
将service文件放置在elasticsearch bin 目录下
mv elasticsearch-servicewrapper-master/service/ bin/
配置bin/service/elasticsearch.conf
vim bin/service/elasticsearch.conf
按需作如下修改
set.default.ES_HOME=/home/elasticsearch-1.5.1 #替换为实际的elasticsearch路径
wrapper.java.command=/usr/lib/jvm/jre-1.7.0-openjdk.x86_64/bin/java #替换为实际的java二进制文件路径
(3). ElasticHQ [非常喜欢]
http://www.elastichq.org/
bin/plugin -i royrusso/elasticsearch-HQ -u file:/home/elasticsearch-1.5.1/royrusso-elasticsearch-HQ-603ae9e.zip
在启动后,可以通过以下方式查看elasticsearch运行情况
http://xxx.xxx.xxx.xxx:8765/_plugin/HQ/
(4) elasticsearch-head [比较喜欢]
https://github.com/mobz/elasticsearch-head
bin/plugin -i mobz/elasticsearch-head -u file:/home/elasticsarch-1.5.1/elasticsearch-head-master.zip
在启动后,可以通过以下方式查看elasticsearch运行情况
http://xxx.xxx.xxx.xxxx:8765/_plugin/head
第四步:启动elasticsearch
bin/service/elasticsearch start|stop|console|install|remove
start 在后台运行elasticsearch
stop 停止elasticsearch
console 在前台运行elasticsearch
install elasticsearch自启动
remove elasticsearch取消自启动
二、基本操作
首先我们批量导入示例数据——莎士比亚全集
(参照http://kibana.logstash.es/content/v3/10-minute-walk-through.html kibana 3指南10分钟入门
wget http://www.elasticsearch.org/guide/en/kibana/3.0/snippets/shakespeare.json
curl -XPUT http://localhost:8765/_bulk --data-binary @shakespeare.json
more shakespeare.json 察看存储内容
{"index":{"_index":"shakespeare","_type":"act","_id":0}}
{"line_id":1,"play_name":"Henry IV","speech_number":"","line_number":"","speaker":"","text_entry":"ACT I"}
{"index":{"_index":"shakespeare","_type":"scene","_id":1}}
接下来我们来通过与熟悉的关系数据库来对比elasticsearch的数据组成
(1)数据组成:元数据+实际数据
相当于察看数据库的模式定义
http localhost:8765/shakespeare/
返回
{
"shakespeare": {
"mappings": {
"act": {
"properties": {
"line_id": {
"type": "long"
},
"line_number": {
"type": "string"
},
"play_name": {
"type": "string"
},
"speaker": {
"type": "string"
},
"speech_number": {
"type": "long"
},
"text_entry": {
"type": "string"
}
}
},
"line": {
"properties": {
"line_id": {
"type": "long"
},
"line_number": {
"type": "string"
},
"play_name": {
"type": "string"
},
"speaker": {
"type": "string"
},
"speech_number": {
"type": "long"
},
"text_entry": {
"type": "string"
}
}
},
"scene": {
"properties": {
"line_id": {
"type": "long"
},
"line_number": {
"type": "string"
},
"play_name": {
"type": "string"
},
"speaker": {
"type": "string"
},
"speech_number": {
"type": "long"
},
"text_entry": {
"type": "string"
}
}
}
},
"settings": {
"index": {
"creation_date": "1429691321987",
"number_of_replicas": "1",
"number_of_shards": "5",
"uuid": "rrCmsKKcSDyLSpLFVnQnbg",
"version": {
"created": "1040299"
}
}
}
}
}
我们用熟悉的关系数据库来进行对比,映射关系如下
elasticsearch RDBS
indices 索引 databases数据库
types 类型 tables表
documents文档 rows行
fields 字段 columns列
示例中,索引名为shakespeare(等同于数据库名为shakespeare)
类型有3个:act, line, scene (等同于表名为act, line, scene)
字段组成(等同于表的结构)
字段名 字段类型
line_id long
line_number string
play_name string
speaker string
speech_number long
text_entry string
(2)简单检索
示例1:通过index+type+文档_id来察看内容
格式:host:port/index_name/type_name/_id
http localhost:8108/shakespeare/line/2
结果如下:
{
"_id": "2",
"_index": "shakespeare",
"_source": {
"line_id": 3,
"line_number": "",
"play_name": "Henry IV",
"speaker": "",
"speech_number": "",
"text_entry": "Enter KING HENRY, LORD JOHN OF LANCASTER, the EARL of WESTMORELAND, SIR WALTER BLUNT, and others"
},
"_type": "line",
"_version": 1,
"found": true
}
elasticsearch的数据由两部分组成:文档元数据(例如_id)与文档数据
名字 说明
_index 类似RDBS的“数据库”概念
_type 类似RDBS的“表”概念
_id 文档的唯一编号
_source 字段里的内容为文档数据(真实存储的数据),我们可以使用如下方法只读取实际数据
http localhost:8108/shakespeare/line/2/_source
结果如下:
{
"line_id": 3,
"line_number": "",
"play_name": "Henry IV",
"speaker": "",
"speech_number": "",
"text_entry": "Enter KING HENRY, LORD JOHN OF LANCASTER, the EARL of WESTMORELAND, SIR WALTER BLUNT, and others"
}
示例2:指定字段field进行搜索,例如搜索play_name字段为Romeo and Juliet
http localhost:8108/shakespeare/_search?q=play_name:"Romeo and Juliet"
结果如下(截取部分):
{
"_shards": {
"failed": 0,
"successful": 5,
"total": 5
},
"hits": {
"hits": [
{
"_id": "86748",
"_index": "shakespeare",
"_score": 3.3792284,
"_source": {
"line_id": 86749,
"line_number": "",
"play_name": "Romeo and Juliet",
"speaker": "JULIET",
"speech_number": 19,
"text_entry": "Exeunt"
},
"_type": "line"
},
(3)复杂搜索
Elasticsearch支持丰富而灵活的查询语言——Query DSL。 在学习之前,我们可以先熟悉一下Lucene查询语法(其实和使用google搜索引擎区别不大)
支持AND,OR,NOT
查询语句"apache AND lucene"的意思是匹配含apache且含lucene的文档。
查询表达式"apache OR lucene"能够匹配包含“apache”的文档,也能匹配包含"lucene"的文档,还能匹配同时包含这两个Term的文档。
查询表达式“lucene NOT elasticsearch”就只能匹配包含lucene但是不含elasticsearch的文档
支持+, -符号
例如:希望搜索到包含关键词lucene,但是不含关键词elasticsearch的文档,可以用如下的查询表达式:"+lucene -elasticsearch"。
支持指定字段名进行搜索(类似RDBS按列名搜索)
例如:查询title域中包含关键词elasticsearch的文档,查询表达式如下:title:elasticsearch
支持通配符
? (匹配单个字符)
* (匹配多个字符)
注意默认的通配符不能是关键词的首字母
支持~整数符号
一个~符号,后面紧跟一个整数,~后面的整数表示短语中可接收的最大的词编辑距离(短语中替换一个词,添加一个词,删除一个词)
"writer~2"能够搜索到含writer和writers的文档。
title:"mastering elasticsearch"~2能够搜匹配title域中含"mastering elasticsearch"的文档与包含"mastering book elasticsearch"的文档
支持^符号进行加权boost设置
一个^符号后面接一个浮点数表示权重。如果权重小于1,就会降低关键词的重要程度。同理,如果权重大于1就会增加关键词的重要程度。默认的加权值为1
支持区间搜索
price:[10.00 TO 15.00查询price域的值在10.00到15.00之间的所有文档。
price:[10.00 TO 15.00}查询price域中价格在10.00(10.00要能够被搜索到)到15.00(15.00不能被搜索到)之间的文档
特殊字符需转义
+, -, &&, || , ! , (,) , { } , [ ] , ^, " , ~, *, ?, : , \, /
更多,Lucene原理 (打分算法,TF-IDF算法一定会在搜索中出境)
我们可以看到elasticsearch支持丰富的数据查询方式,结果展示方式(按什么方式来排序结果,使用什么图形来展示统计结果)
(1)关键词查询term
(2)短语查询phrase
(3)区间range
(4)布尔Boolean
(5)模糊fuzzy
(6)跨度span
(7)通配符wildcard
(8)地理位置spatial
(9) 统计aggregation ——这个功能非常非常赞,比如说生成各种统计图表
(10)prospective search
搜索语句支持通过URI提交(上面的例子演示的_search?q= 注意,使用这种方式的要遵循url编码,官方参考) ,也支持通过request body提交,简直就是HTTP RESTFULL最佳实践,官方参考
我们用熟悉的SQL语句来对比
实例1:
curl -XPOST 'http://localhost:8108/shakespeare/line/_search?pretty' -d '
{
"query":{ "match_all": {} },
"sort": {"line_id": {"order": "desc" }},
"size": 1,
"from": 10
}'
等同于
use shakespeare;
select *
from line
order by line_id desc
limit 10,1
实例2:
curl -XPOST 'http://localhost:8108/shakespeare/line/_search?pretty' -d '
{
"query":{
"bool":{
"must":[
{"match_phrase": {"text_entry":"question"}},
{"match_phrase": {"text_entry":"not to be"}}
]
}
}
}'
结果
"took" : 253,
"timed_out" : false,
"_shards" : {
"total" : 3,
"successful" : 3,
"failed" : 0
},
"hits" : {
"total" : 2,
"max_score" : 4.0433946,
"hits" : [ {
"_index" : "shakespeare",
"_type" : "line",
"_id" : "34229",
"_score" : 4.0433946,
"_source":{"line_id":34230,"play_name":"Hamlet","speech_number":19,"line_number":"3.1.64","speaker":"HAMLET","text_entry":"To be, or not to be: that is the question:"}
}, {
"_index" : "shakespeare",
"_type" : "line",
"_id" : "1397",
"_score" : 4.0004296,
"_source":{"line_id":1398,"play_name":"Henry IV","speech_number":152,"line_number":"2.4.392","speaker":"FALSTAFF","text_entry":"blackberries? a question not to be asked. Shall"}
} ]
}
}
等同于
use shakespeare;
select *
from line
where text_entry like "%question%" and text_entry like "%not to be%"
Search APIs
Match Query APIs