Redis Enterprise 的混合语义搜索功能 redis elasticsearch 组合

转载

数据分析大师 2023-10-13 22:40:28

文章标签 elasticsearch java bc 字段 文章分类 Redis 数据库

一、springBoot集成Es

1、引入jar包依赖

<dependency>
      <groupId>org.springframework.boot</groupId>
      <artifactId>spring-boot-starter-data-elasticsearch</artifactId>
</dependency>

<dependency>
       <groupId>org.elasticsearch.plugin</groupId>
       <artifactId>transport-netty4-client</artifactId>
       <version>6.2.3</version>
</dependency>

2、配置连接属性

spring: 
    data:
        elasticsearch:
            repositories:
                enabled: true
            cluster-name: elasticsearch
            cluster-nodes: 127.0.0.1:9300

3、springboot版本为2.1.9.RELEASE，与es版本对应关系如图：

Redis Enterprise 的混合语义搜索功能 redis elasticsearch 组合_java

4、启动项目

项目启动报错：Error creating bean with name 'elasticsearchClient', AvailableProcessors is already set to [4] 百度上解释是es和 redis产生冲突

解决方式：在启动类中加上

@SpringBootApplication
public class HumanPersonalApplication {

    public static void main(String[] args) {
        System.setProperty("es.set.netty.runtime.available.processors","false");
        SpringApplication.run(HumanPersonalApplication.class, args);
    }

}

在测试类中加上

public EsTest() {
        System.setProperty("es.set.netty.runtime.available.processors","false");
    }

二、下载对应版本的Elasticsearch、Logstash、kibana进行安装

下载路径：https://www.elastic.co/cn/downloads/

三、同步数据两种方式：

1、使用业务代码实现同步

1.1 在业务层进行数据同步，将查询的结果进行封装成对象同步到Es中

@Test
    public void same(){
        //创建索引
        elasticsearchTemplate.createIndex(Resume.class);
        List<Resume> list = enterpriseBidStaffMapper.selectAll();
        for (Resume r:list) {
            List<EducationExperiences> educationExperiences = r.getEducationExperiences();
            if(educationExperiences!=null && educationExperiences.size()>0){
                for (EducationExperiences e:educationExperiences) {
                    if(e.getEducation()!=null){
                        e.setEducationo(Transformation.getEducation(e.getEducation()));
                    }
                }
            }
        }
        //批量同步数据
        esRespostory.saveAll(list);
        System.out.println(list.size());
    }

1.2 关于LocalDateTime反序列化问题处理方案：

/**
    * 创建时间
    */
    @DateTimeFormat(pattern = "yyyy-MM-dd\'T\'HH:mm:ss.SSS")
    @JsonSerialize(using = LocalDateTimeSerializer.class)
    @JsonDeserialize(using = LocalDateTimeDeserializer.class)
    private LocalDateTime createTime;

1.3 对于分词字段与不分词字段设置

/**
	 * FieldType.Text 定义该字段需要分词
	 */
	@Field(type = FieldType.Text,analyzer = "ik_max_word")
	private String hopeOccupationName;

        /**
	 * FieldType.Keyword定义字段不需要分词
	 */
	@Field(type = FieldType.Keyword)
	private String jobNature;

1.4 对于es分页查询

Integer pageNo = Integer.valueOf(maps.get("page").toString());
        Integer pageSize = Integer.valueOf(maps.get("limit").toString());
        //pageNo-1当前页
        //每页记录数
        //Sort.by(Sort.Direction.DESC,"排序字段")
        PageRequest pageRequest = PageRequest.of(pageNo-1, pageSize,   Sort.by(Sort.Direction.DESC,"id"));
        org.springframework.data.domain.Page<Resume> page = esRespostory.search(boolQueryBuilder,pageRequest);
        List<Resume> resumes = page.getContent();
        Map<String, Object> result = new HashMap<>();
        result.put("records" ,resumes);
        result.put("total", page.getTotalElements());
        result.put("size",pageSize);
        result.put("current",pageNo);
        int totalElements = (int)page.getTotalElements();
        result.put("pages",totalElements%pageSize==0?totalElements/pageSize:totalElements/pageSize+1);

2 、使用Logstash进行数据增量同步

1、修改配置文件

# Sample Logstash configuration for creating a simple
# Beats -> Logstash -> Elasticsearch pipeline.

input {

  stdin {}
  
  jdbc {
    jdbc_driver_library => "E:\\Elasticsearch\\logstash-6.4.3\\mysql-connector-java-5.1.47.jar"
    jdbc_driver_class => "com.mysql.jdbc.Driver"
    jdbc_connection_string => "jdbc:mysql://ip:port/sipaote_human_test?useUnicode=true&characterEncoding=utf8&allowMultiQueries=true"
    jdbc_user => "name"
    jdbc_password => "pwd"
    jdbc_paging_enabled => true
    #是否记录上次执行结果，如果为真，将会把上次执行到的tracking_column字段的值记录下来，保存到last_run_metadata_path指定的文件中
    record_last_run => true
    #是否需要记录某个column的值
    use_column_value => true
    #如果use_column_value的值为true,需配置此参数，track的数据库column名，该column必须是递增的
    tracking_column => "updateTime"
    tracking_column_type => "timestamp"
    last_run_metadata_path => "E:\\Elasticsearch\\logstash-6.4.3\\.logstash_jdbc_last_run"
    lowercase_column_names => false
    #处理中文乱码问题
    codec => plain { charset => "UTF-8"}
    # 索引类型
    #type => "bid"
    clean_run => false
    schedule => "*/5 * * * * *"
    statement_filepath => "E:\\Elasticsearch\\logstash-6.4.3\\my.sql"
    #statement => ""
  }
}

filter {
     aggregate {
        task_id => "%{bsid}"
        code => "
            map['bsid'] = event.get('bsid')
            map['hopeCityName'] = event.get('hopeCityName')
            map['hopeIndustryName'] = event.get('hopeIndustryName')
            map['hopeOccupationName'] = event.get('hopeOccupationName')
			map['jobNature'] = event.get('jobNature')
			map['name'] = event.get('name')
			map['tel'] = event.get('tel')
			map['createTime'] = event.get('createTime')
			map['sex'] = event.get('sex')
			map['createUser'] = event.get('createUser')
			map['enterpriseId'] = event.get('enterpriseId')
			map['headhuntingName'] = event.get('headhuntingName')
			map['isBlack'] = event.get('isBlack')
			#使用map['w_list']去除重复
			map['w_list'] ||=[]
			map['workExperience'] ||=[]
			if(event.get('wid')!=nil)
				if !(map['w_list'].include? event.get('wid'))
					map['w_list'] << event.get('wid')
					map['workExperience'] << {
						'wid' => event.get('wid'),
						'industryName' => event.get('industryName'),
						'skillsDescribe' => event.get('skillsDescribe')
					}
				end
			end
			map['e_list'] ||=[]
			map['educationExperiences'] ||=[]
			if(event.get('eid')!=nil)
				if !(map['e_list'].include? event.get('eid'))
					map['e_list'] << event.get('eid')
					map['educationExperiences'] << {
						'eid' => event.get('eid'),
						'schoolName' => event.get('schoolName'),
						'speciality' => event.get('speciality')
					}
				end
			end
			map['s_list'] ||=[]
			map['skill'] ||=[]
			if(event.get('sid')!=nil)
				if !(map['s_list'].include? event.get('sid'))
					map['s_list'] << event.get('sid')
					map['skill'] << {
						'sid' => event.get('sid'),
						'skillName' => event.get('skillName')
					}
				end
			end
			map['c_list'] ||=[]
			map['certificate'] ||=[]
			if(event.get('cid')!=nil)
				if !(map['c_list'].include? event.get('cid'))
					map['c_list'] << event.get('cid')
					map['certificate'] << {
						'cid' => event.get('cid'),
						'certificateName' => event.get('certificateName')
					}
				end
			end
            event.cancel()
        "
        
        push_previous_map_as_event => true
        timeout => 5
    }
    json {
        source => "message"
        remove_field => ["message"]
        #remove_field => ["message", "type", "@timestamp", "@version"]
    }
    mutate  {
        #将不需要的JSON字段过滤，且不会被存入 ES 中
        remove_field => ["tags", "@timestamp", "@version"]
    }
}



output {

 #if[type]=="bid"{
	elasticsearch {
		hosts => ["127.0.0.1:9200"]
		index => "human"
		document_type => "resumans"
		document_id => "%{bsid}"
    }
  
    stdout{
		codec => json
    }
 #}
  
}

2.2 多表关联查询语句

SELECT
	bs.id bsid,
	bs.hope_city_name hopeCityName,
	bs.hope_industry_name hopeIndustryName,
	bs.hope_occupation_name hopeOccupationName,
	bs.job_nature jobNature,
	bs.job_status jobStatus,
	bs.`name`,
	bs.tel,
	bs.create_time createTime,
	bs.sex,
	bs.create_user createUser,
	IF(
		IFNULL(bs.update_time,'1997-1-1 00:00:00')>IFNULL(w.update_time,'1997-1-1 00:00:00'),
		IF(
			IFNULL(bs.update_time,'1997-1-1 00:00:00')>IFNULL(e.update_time,'1997-1-1 00:00:00'),
			IF(
				IFNULL(bs.update_time,'1997-1-1 00:00:00')>IFNULL(c.update_time,'1997-1-1 00:00:00'),
				IF(
					IFNULL(bs.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(bs.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),bs.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				),
				IF(IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),c.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				)
			),
			IF(
				IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(c.update_time,'1997-1-1 00:00:00'),
				IF(
					IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),e.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				),
				IF(
					IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),c.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				)
			)
		),
		IF(
			IFNULL(w.update_time,'1997-1-1 00:00:00')>IFNULL(e.update_time,'1997-1-1 00:00:00'),
			IF(
				IFNULL(w.update_time,'1997-1-1 00:00:00')>IFNULL(c.update_time,'1997-1-1 00:00:00'),
				IF(
					IFNULL(w.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(w.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),w.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				),
				IF(
					IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),c.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				)
			),
			IF(
				IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(c.update_time,'1997-1-1 00:00:00'),
				IF(
					IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),e.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				),
				IF(
					IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),c.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				)
			)
		)
	) updateTime,
	w.id wid,
	w.industry_name industryName,
	w.skills_describe skillsDescribe,
	e.id eid,
	e.school_name schoolName,
	e.speciality speciality,
	e.education education,
	c.id cid,
	c.certificate_name certificateName,
	s.id sid,
	s.skill_name skillName,
	u.`name` headhuntingName,
	IF(k.id is NULL,0,1) isBlack
FROM s_enterprise_bid_staff bs
LEFT JOIN s_work_experience w ON bs.id=w.bid_staff_id
LEFT JOIN s_education_experience e ON bs.id=e.bid_staff_id
LEFT JOIN s_certificate c ON bs.id=c.bid_staff_id
LEFT JOIN s_skill s ON bs.id=s.bid_staff_id
LEFT JOIN s_enterprise_user u ON bs.bid_enterprise_id=u.id
LEFT JOIN s_blacklist k ON k.bid_staff_id=bs.id AND is_delete=0
where 
bs.id in (
	SELECT
		bs.id bsid
	FROM s_enterprise_bid_staff bs
	LEFT JOIN s_work_experience w ON bs.id=w.bid_staff_id
	LEFT JOIN s_education_experience e ON bs.id=e.bid_staff_id
	LEFT JOIN s_certificate c ON bs.id=c.bid_staff_id
	LEFT JOIN s_skill s ON bs.id=s.bid_staff_id
	LEFT JOIN s_enterprise_user u ON bs.bid_enterprise_id=u.id
	LEFT JOIN s_blacklist k ON k.bid_staff_id=bs.id
	where 
	IF(
		IFNULL(bs.update_time,'1997-1-1 00:00:00')>IFNULL(w.update_time,'1997-1-1 00:00:00'),
		IF(
			IFNULL(bs.update_time,'1997-1-1 00:00:00')>IFNULL(e.update_time,'1997-1-1 00:00:00'),
			IF(
				IFNULL(bs.update_time,'1997-1-1 00:00:00')>IFNULL(c.update_time,'1997-1-1 00:00:00'),
				IF(
					IFNULL(bs.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(bs.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),bs.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				),
				IF(IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),c.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				)
			),
			IF(
				IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(c.update_time,'1997-1-1 00:00:00'),
				IF(
					IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),e.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				),
				IF(
					IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),c.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				)
			)
		),
		IF(
			IFNULL(w.update_time,'1997-1-1 00:00:00')>IFNULL(e.update_time,'1997-1-1 00:00:00'),
			IF(
				IFNULL(w.update_time,'1997-1-1 00:00:00')>IFNULL(c.update_time,'1997-1-1 00:00:00'),
				IF(
					IFNULL(w.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(w.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),w.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				),
				IF(
					IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),c.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				)
			),
			IF(
				IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(c.update_time,'1997-1-1 00:00:00'),
				IF(
					IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(e.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),e.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				),
				IF(
					IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(s.update_time,'1997-1-1 00:00:00'),
					IF(IFNULL(c.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),c.update_time,k.update_time),
					IF(IFNULL(s.update_time,'1997-1-1 00:00:00')>IFNULL(k.update_time,'1997-1-1 00:00:00'),s.update_time,k.update_time)
				)
			)
		)
	)> :sql_last_value
)
ORDER BY bs.id DESC

2.3 遇到的问题与解决方案：

2.3.1 logstash同步nested嵌套数据类型：创建对应的mapping 映射关系

PUT /human
{
  "mappings": {
    "resumans": {
      "properties": {
        "bsid":{
          "type": "integer"
        },
        "hopeCityName":{
          "type": "text",
          "analyzer": "ik_max_word",
          "search_analyzer": "ik_max_word"
        },
        "hopeIndustryName":{
          "type": "text",
          "analyzer": "ik_max_word",
          "search_analyzer": "ik_max_word"

        },
        "hopeOccupationName":{
          "type": "text"
        },
        "jobNature":{
          "type": "text"
        },
        "name":{
          "type": "keyword"
        },
        "tel":{
          "type": "keyword"
        },
        "createTime":{
          "type": "date"
        },
        "sex":{
          "type": "integer"
        },
        "createUser":{
          "type": "integer"
        },
        "enterpriseId":{
          "type": "integer"
        },
        "headhuntingName":{
          "type": "keyword"
        },
        "isBlack":{
          "type": "integer"
        },
        "workExperience":{
           "type": "nested", 
           "properties": {
             "wid":{
               "type":"integer"
             },
             "industryName":{
               "type":"text"
             },
             "skillsDescribe":{
               "type":"text"
             }
           }
        },
        "educationExperiences":{
          "type": "nested",
          "properties":{
            "eid":{
              "type":"integer"
            },
           "schoolName":{
              "type":"text",
              "analyzer": "ik_max_word",
              "search_analyzer": "ik_max_word"
            },
            "speciality":{
              "type":"text",
              "analyzer": "ik_max_word",
              "search_analyzer": "ik_max_word"
            }
          }
        },
        "skill":{
          "type": "nested",
          "properties":{
            "sid":{
            "type":"integer"
          },
          "skillName":{
            "type":"text",
            "analyzer": "ik_max_word",
            "search_analyzer": "ik_max_word"
          }
          }
        },
        "certificate":{
          "type": "nested",
          "properties":{
            "cid":{
            "type":"integer"
          },
          "certificateName":{
            "type":"text",
            "analyzer": "ik_max_word",
            "search_analyzer": "ik_max_word"
          }
          }
        }
      }
    }
  }
}

2.3.2 logstash同步数据时会出现子对象数组中的对象数据出现重复：

Redis Enterprise 的混合语义搜索功能 redis elasticsearch 组合_elasticsearch_02

解决方法：

#使用map['w_list']去除重复
    map['w_list'] ||=[]
    map['workExperience'] ||=[]
    if(event.get('wid')!=nil)
        if !(map['w_list'].include? event.get('wid'))
            map['w_list'] << event.get('wid')
            map['workExperience'] << {
                'wid' => event.get('wid'),
                'industryName' => event.get('industryName'),
                'skillsDescribe' => event.get('skillsDescribe')
            }
        end
    end

2.3.3 logstash同步时少同步一条数据，在停止logstash服务时才进行同步

在filter 聚合配置中添加：timeout => 5

filter aggregate 创建中 event map 并不知道这次事件是不是应该结束，也就是它也不知道到那一条才是最后一条, 因此设置一个 timeout 告诉它这个时间执行多少秒就结束继续执行第二个。但这样并不是很严谨，因为你也不确定你的 event map 到底要执行多久。最好的方式是我们应该给定一个 task end 的条件

本文章为转载内容，我们尊重原作者对文章享有的著作权。如有内容错误或侵权问题，欢迎原作者联系我们进行内容更正或删除文章。

上一篇：MySQL学生表怎么和班级表连接 mysql创建学生数据表

下一篇：python kafka 生产者指定分区 kafka生产者api

提问和评论都可以，用心的回复会被更多人看到评论

发布评论

相关文章

官方博客	全部文章	热门标签	班级博客
了解我们	网站地图	意见反馈

鸿蒙开发者社区	51CTO学堂
51CTO	软考资讯