Spring Elasticsearch(Neumann)

简介

从文档 Spring Data Elasticsearch - Reference Documentation 可以查看到Spring 与 Elasticsearch之间的版本关系。

这里由于项目的需要，我们选取的是 Neumann版本，对应的是

• Spring Data Elasticsearch 4.0.x
• Elasticsearch 7.6.2
• Spring Framework 5.2.12
• Spring Boot 2.3.x

官方文档 Spring Data Neumann goes GA，注意需要使用的是这个文档。

Elasticsearch 和 数据库 对比

测试方法

我们使用如下测试代码作为测试基础

package com.example.demo;

import org.junit.jupiter.api.Test;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.boot.test.context.SpringBootTest;

@SpringBootTest
public class ElasticsearchTests {

    @Autowired
    private ElasticsearchRestTemplate elasticsearchRestTemplate;

    @Test
    void testQuery() {
        // 测试代码
    }
}

只需要修改testQuery内代码，就可以完成针对不同代码的测试。

查询所有文档

查询一个索引中的所有数据。类似于查询语句

select * from metric_server;

接下来是获取Elasticsearch数据

@Data
class TestObject {
    private String name;
    private String region;
    private String resourceId;
}

@Test
void testQuery() {
    Query query = new NativeSearchQueryBuilder()
            .withQuery(matchAllQuery())
            .build();
    
    IndexCoordinates indexCoordinates = IndexCoordinates.of("metric_server");
    
    List<TestObject> objs =  elasticsearchRestTemplate.queryForList(query, TestObject.class, indexCoordinates);

    for ( int i = 0 ; i < objs.size() ;i++) {
        System.out.println(objs.get(i));
    }
}

根据条件简单查询

加上条件再做查询,类似于以下

select * from metric_server where name = 'agenttest';

代码也比较简单，使用termQuery来匹配

@Test
void testQuery() {
    TermQueryBuilder termQueryBuilder = QueryBuilders.termQuery("name", "agenttest");

    Query query = new NativeSearchQueryBuilder()
            .withQuery(termQueryBuilder)
            .build();
    
    IndexCoordinates indexCoordinates = IndexCoordinates.of("metric_server");
    
    List<TestObject> objs =  elasticsearchRestTemplate.queryForList(query, TestObject.class, indexCoordinates);

    for ( int i = 0 ; i < objs.size() ;i++) {
        System.out.println(objs.get(i));
    }
}

各匹配类型说明

根据条件做复杂查询

做一个比较复杂的查询

select * from metric_server where name = 'agenttest' and timestamp > '2021-02-10 00:00:00'

同样的，我们可以使用下面代码做测试

@Test
void testQuery() {
    long sinceTime = System.currentTimeMillis() - 1000L*60*60;;

    BoolQueryBuilder boolQueryBuilder = QueryBuilders.boolQuery();

    boolQueryBuilder.must(termQuery("name", "agenttest"));
    boolQueryBuilder.must(QueryBuilders.rangeQuery("timestamp").gte(sinceTime));
    // boolQueryBuilder.filter(QueryBuilders.rangeQuery("timestamp").gte(sinceTime));
    // 同上

    Query query = new NativeSearchQueryBuilder()
            .withQuery(boolQueryBuilder)
            .build();

    IndexCoordinates indexCoordinates = IndexCoordinates.of("metric_server");

    List<TestObject> objs =  elasticsearchRestTemplate.queryForList(query, TestObject.class, indexCoordinates);

    for ( int i = 0 ; i < objs.size() ;i++) {
        System.out.println(objs.get(i));
    }
}

运行如下图

其中filter和must在这里是一致的，但是他们在ES里还是有差别的，详细参考

• ElasticSearch中must和filter的区别
• Es检索 must与filter区别

查询并排序

查询并测试

select * from metric_server where name = 'agenttest' order by timestamp desc;

使用下面代码测试

@Data
class TestObject {
    private String name;
    private String region;
    private String resourceId;
    private String timestamp;
}

@Test
void testQuery() {

    TermQueryBuilder termQuery = QueryBuilders.termQuery("name", "agenttest");

    SortBuilder sortBuilder = SortBuilders.fieldSort("timestamp").order(SortOrder.DESC);

    Query query = new NativeSearchQueryBuilder()
            .withQuery(termQuery)
            .withSort(sortBuilder)
            .build();

    IndexCoordinates indexCoordinates = IndexCoordinates.of("metric_server");

    List<TestObject> objs =  elasticsearchRestTemplate.queryForList(query, TestObject.class, indexCoordinates);

    System.out.println(objs.size());

    for ( int i = 0 ; i < objs.size() ;i++) {
        System.out.println(objs.get(i));
    }
}

查看结果

参考

• springboot中ElasticSearch入门与进阶：组合查询、聚合查询
• Spring Data Elasticsearch使用
• 【必看】SpringBoot整合Spring Data Elasticsearch
• ElasticSearch AggregationBuilders java api常用聚会查询
• ElasticSearch-AggregationBuilders查询遇到的坑
• Elasticsearch Aggregation 笔记
• Spring Data Elasticsearch