Elasticsearch笔记
资料
官网: http://www.elasticsearch.org
.Net驱动: http://nest.azurewebsites.net/
教程: http://www.jianshu.com/p/05cff717563c
三级: 索引/类型/ID或操作符
对应: 数据库/表/ID
配置
启动服务: /es/elasticsearch -d
添加认证: http://segmentfault.com/a/1190000002803609
用户名: admin , admin_pw
关闭: curl -XPOST http://localhost:9200/_cluster/nodes/_shutdown
关闭某一个索引: curl -XPOST http://localhost:9200/_cluster/nodes/[某一个索引]/_shutdown
查看索引: curl -XGet http://localhost:9200/_cluster/nodes
插件
Head
elasticsearch/bin/plugin -install mobz/elasticsearch-head
http://localhost:9200/_plugin/head/
Docker 安装 head
高版本的需要用新的方式安装 head
https://www.cnblogs.com/aubin/p/8018081.html
分词
IK官网:https://github.com/medcl/elasticsearch-analysis-ik
安装IK:bin/plugin -install medcl/elasticsearch-analysis-ik
安装分词需要 Jdk7,Maven,Git,在最新的ES(1.7.1)上,网上的文章大部分都是错的。正确方法:
http://my.oschina.net/xiaohui249/blog/232784
http://www.360doc.com/content/15/0114/13/16070877_440673208.shtml
如果这两篇文章存在有冲突的地方,使用后者。
步骤:
1.安装Jdk7,然后,再运行: set JAVA_HOME=C:\Program Files\Java\jdk1.7.0_80\bin
注意:安装的Jdk版本一定要和服务器使用的JRE版本一致。否则ES启动后,不可使用。
2.安装 Maven,Git
3.进入 D:\ik ,运行: clone https://github.com/medcl/elasticsearch-analysis-ik
4.进入 D:\ik\elasticsearch-analysis-ik
5.运行:mvn clean package
6.生成的 jar 在 D:\ik\elasticsearch-analysis-ik\target\releases\elasticsearch-analysis-ik-1.4.0.zip , 其它的 jar 全是扯蛋。把它解出来。
附上链接: 百度网盘下载:elasticsearch-analysis-ik-1.4.0.zip
7. 把 解压缩的jar 拷贝到: ES_HOME/plugins/analysis-ik
8. 将下载目录中的ik目录拷贝到ES_HOME/config目录下面。
9.打开ES_HOME/config/elasticsearch.yml文件,在文件最后加入如下内容
index: analysis: analyzer: ik: alias: [ik_analyzer] type: org.elasticsearch.index.analysis.IkAnalyzerProvider ik_max_word: type: ik use_smart: false ik_smart: type: ik use_smart: true
或:
index.analysis.analyzer.default.type: ik
10. 重新启动ES
11. 建一个索引,测试:
curl -XPut 'http://localhost:9200/index/_analyze?analyzer=ik&pretty=true' -d '
{
"text":"世界如此之大"
}'
Maven
安装教程:http://jingyan.baidu.com/article/1709ad808ad49f4634c4f00d.html
官网:http://maven.apache.org/download.cgi
Git
下载:http://www.git-scm.com/download/win
过程
1. 创建索引: Put ~/index
2. 获取所有Index: Get ~/_aliases
3. 创建别名: Put ~/<Index>/_alias/<Index_Alias>
4. 创建模板: Put ~/<index>/_mapping/<type> {type: {properties: { "列": { type:"string" }, "列2" : { type: { Id: { type: "int"} , Name: { type: "string"} } } }}}
对于《列2》来说,是一个对象类型,它可以是对象,也可以是对象数组。
更复杂的实例:
{ "hr": { "properties": { "ResumeName": { "type": "string" }, "Source": { "type": "integer" }, "PersonInfo": { "properties": { "Name": { "type": "string", "analyzer": "ik", "include_in_all": true }, "Sex": { "type": "integer", "analyzer": "ik", "include_in_all": true }, "BirthDay": { "type": "date", "analyzer": "ik", "include_in_all": true }, "Mobile": { "type": "string" }, "StartWorkAt": { "type": "date", "analyzer": "ik", "include_in_all": true }, "LiveCity": { "type": "string", "analyzer": "ik", "include_in_all": true }, "JobWantedStatus": { "type": "integer", "analyzer": "ik", "include_in_all": true }, "NativePlace": { "type": "string" }, "MonthlySalary": { "type": "integer" }, "YearSalary": { "type": "integer" }, "National": { "type": "integer" }, "IDCarNumber": { "type": "string" }, "Height": { "type": "integer" }, "MaritalStatus": { "type": "integer" }, "WxNumber": { "type": "string" }, "QqNumber": { "type": "integer" }, "LiveAddress": { "type": "string" }, "HomePage": { "type": "string" } } }, "Intention": { "properties": { "Keyword": { "type": "string" }, "SelfEvaluation": { "type": "string" }, "ExpectWorkIn": { "type": "string", "analyzer": "ik", "include_in_all": true }, "IndustryID": { "type": "integer" }, "IndustryName": { "type": "string" }, "PositionID": { "type": "integer" }, "PositionName": { "type": "string" }, "Nature": { "type": "integer" }, "ExpectSalary": { "type": "integer", "analyzer": "ik", "include_in_all": true }, "ArriveAt": { "type": "date" } } }, "Education": { "properties": { "StartAt": { "type": "date" }, "EndAt": { "type": "date" }, "School": { "type": "string" }, "Major": { "type": "string", "analyzer": "ik", "include_in_all": true }, "Education": { "type": "integer", "analyzer": "ik", "include_in_all": true }, "Remark": { "type": "string" }, "IsForeign": { "type": "boolean" } } }, "Experience": { "properties": { "StartAt": { "type": "date" }, "EndAt": { "type": "date" }, "Corporation": { "type": "string" }, "IndustryID": { "type": "integer" }, "IndustryName": { "type": "string" }, "Size": { "type": "integer" }, "Nature": { "type": "integer" }, "PositionID": { "type": "integer" }, "PositionName": { "type": "string" }, "Remark": { "type": "string" }, "WorkTimeType": { "type": "integer" } } }, "Train": { "properties": { "StartAt": { "type": "date" }, "EndAt": { "type": "date" }, "TrainingInstitution": { "type": "string" }, "TrainingPlace": { "type": "string" }, "TrainingCourse": { "type": "string" }, "Certificate": { "type": "string" }, "Remark": { "type": "string" } } }, "Language": { "properties": { "LanguageType": { "type": "integer" }, "Degree": { "type": "integer" }, "Level": { "type": "string" } } }, "AdditionalInfo": { "properties": { "Subject": { "type": "string" }, "Content": { "type": "string" } } }, "Certificate": { "properties": { "Name": { "type": "string" }, "GetAt": { "type": "date" }, "Score": { "type": "string" } } }, "Project": { "properties": { "StartAt": { "type": "date" }, "EndAt": { "type": "date" }, "Name": { "type": "string", "analyzer": "ik", "include_in_all": true }, "Development": { "type": "string" }, "Remark": { "type": "string" }, "MyDuty": { "type": "string" } } }, "ItSkill": { "properties": { "Name": { "type": "string" }, "UseTime": { "type": "string" }, "Degree": { "type": "integer" } } }, "TxtContent": { "type": "string", "analyzer": "ik", "include_in_all": true }, "OtherContent": { "type": "string", "analyzer": "ik", "include_in_all": true } } } }
5. 获取映射信息: Get ~/<index>/_mapping/<type>
6. 查看所有的索引及类型信息: Get ~/_mapping
7. 查看状态:
http://blog.chinaunix.net/uid-532511-id-4854331.html
健康状态: curl -XGET 'http://localhost:9200/_cluster/health?pretty=true'
8. 集群状态:
curl -XGET 'http://localhost:9200/_cluster/state?pretty'
curl -XGET 'http://localhost:9200/_cluster/stats?human&pretty'
9. 节点状态:
curl -XGET 'http://localhost:9200/_nodes/stats?pretty'
curl -XGET 'http://localhost:9200/_nodes/stats/os,process?pretty'
curl -XGET 'http://localhost:9200/_nodes?pretty'
curl -XGET 'http://localhost:9200/_nodes/process?pretty'
curl -XGET 'http://localhost:9200/_nodes/os?pretty'
curl -XGET 'http://localhost:9200/_nodes/settings?pretty'
10. 本机节点状态:
curl -XGET 'localhost:9200/_nodes/_local?pretty'
curl -XGET 'localhost:9200/_nodes/_local/network?pretty'
查询也可以针对子节点进行:
curl -XGET 'http://localhost:9200/_nodes/子节点?pretty'
Mapping
官网类型定义:https://www.elastic.co/guide/en/elasticsearch/guide/current/mapping-intro.html
官网复合类型: https://www.elastic.co/guide/en/elasticsearch/guide/current/complex-core-fields.html
复合类型是指,值是:数组,对象。
任何值,都可以按定义的数据类型保存为一个值,或一个该值的数据。
知识点
springboot 两个端口,两种驱动
建议使用 9200 的 Rest 方式。
http://www.jb51.net/article/127390.htm
查询
http://www.cnblogs.com/zhangchenliang/p/4195406.html
排序 function_score函数
官方:https://www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-function-score-query.html
文章: http://blog.csdn.net/dm_vincent/article/details/42201789
权重
字段后使用 ^2 表示该字段权限是2 如:
query:"title:中国^2+title:北京-content:美国" 表示:title中包含中国的部分权重是2 ,同时 title中包含北京,同时, content中不包含 美国。
返回列
_source 返回该节点及子节点
"_source": { "Id": "8a7229609bd841fcb7ed2ffc68a5a01f", "PersonInfo": { "Name": "马少明" } },
fields 只能返回叶子节点
"fields": { "PersonInfo.Name": [ "马少明" ], "Id": [ "8a7229609bd841fcb7ed2ffc68a5a01f" ] },
问题
1. 提交的Json,不能带有如下属性 : id , _id , 但可以是: Id ,ID !
ES里的系统字段包含: _index,_type,_id,_score
2. 提交的Json,第一次提交,会建立一个模板,如果第二次提交的数据与第一次提交的数据在格式有冲突,则不能创建。
典型的例子是,第一次提交的数据是,有一个对象数组,但值是空的。 第二次提交的对象数据有值,则不能提交。
需要修改数据模板。
修改索引:
http://www.cnblogs.com/Creator/p/3722408.html
http://www.cnblogs.com/tianjixiaoying/p/4424076.html
3.健康值红了:
http://localhost:9200/wy/_optimize
4. 使用 function_score 的文章: http://blog.csdn.net/dm_vincent/article/details/42201721 , 它里面写的: Get _search 是错的,应该是 Post _search
Post ~/_search or ~/wy/_search or ~/wy/hr/_search
{ "query": { "function_score": { "filter": { "term": { "BirthYear": "1980" } }, "functions": [ { "filter": { "term": { "NowCity": "北京" } }, "weight": 1 } ], "score_mode": "sum" } } }
使用SQL进行ES查询
其实ES官方早就应该实现这个功能: http://www.open-open.com/lib/view/open1447747736056.html
作者:NewSea 出处:http://newsea.cnblogs.com/
QQ,MSN:iamnewsea@hotmail.com 如无特别标记说明,均为NewSea原创,版权私有,翻载必纠。欢迎交流,转载,但要在页面明显位置给出原文连接。谢谢。 |