如何通过Spring Data Elasticsearch更新现有索引映射且不丢失数据？

首先你遇到的mapper_parsing_exception是因为期望的映射结构存在语法错误：在_doc节点下重复嵌套了mappings字段，这不符合Elasticsearch的映射规范，是导致解析失败的核心原因。先修正目标映射结构，再用Spring Data Elasticsearch实现增量更新。

第一步：修正目标映射结构 #

正确的映射应该去掉_doc下多余的mappings节点，结构如下：

{
  "_doc": {
    "dynamic_templates": [
      {
        "strings": {
          "match_mapping_type": "string",
          "mapping": {
            "type": "keyword"
          }
        }
      },
      {
        "dates": {
          "match_mapping_type": "date",
          "mapping": {
            "type": "date",
            "format": "yyyy-MM-dd HH:mm:ss"
          }
        }
      }
    ],
    "properties": {
      "_class": {
        "type": "text",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 420
          }
        }
      },
      "date": {
        "type": "long"
      },
      "favorite_color": {
        "type": "keyword"
      },
      "is_active": {
        "type": "boolean"
      },
      "number_float": {
        "type": "float"
      },
      "number_long": {
        "type": "long"
      },
      "taskType": {
        "type": "integer"
      },
      "updateTime": {
        "type": "date",
        "format": "yyyy-MM-dd HH:mm:ss"
      },
      "updatedDate": {
        "type": "date",
        "format": "yyyy-MM-dd HH:mm:ss"
      },
      "workType": {
        "type": "integer"
      }
    }
  }
}

第二步：用Spring Data Elasticsearch实现映射更新 #

可以通过两种方式实现，均不会丢失现有数据：

方式一：实体类+Repository自动同步映射 #

1. 创建对应目标映射的实体类

import org.springframework.data.elasticsearch.annotations.Document;
import org.springframework.data.elasticsearch.annotations.DynamicTemplate;
import org.springframework.data.elasticsearch.annotations.Field;
import org.springframework.data.elasticsearch.annotations.FieldType;

import java.util.Date;

@Document(indexName = "my-index")
public class MyIndexEntity {
    // 原有字段
    @Field(type = FieldType.Text, fields = {@Field(type = FieldType.Keyword, ignoreAbove = 420)})
    private String _class;

    @Field(type = FieldType.Integer)
    private Integer taskType;

    @Field(type = FieldType.Date, format = "yyyy-MM-dd HH:mm:ss")
    private Date updateTime;

    @Field(type = FieldType.Integer)
    private Integer workType;

    // 新增字段
    @Field(type = FieldType.Long)
    private Long date;

    @Field(type = FieldType.Keyword)
    private String favorite_color;

    @Field(type = FieldType.Boolean)
    private Boolean is_active;

    @Field(type = FieldType.Float)
    private Float number_float;

    @Field(type = FieldType.Long)
    private Long number_long;

    @Field(type = FieldType.Date, format = "yyyy-MM-dd HH:mm:ss")
    private Date updatedDate;

    // 配置动态模板（部分Spring Data ES版本需配合外部mapping文件，注解支持以实际版本为准）
    @DynamicTemplates({
        @DynamicTemplate(
            name = "strings",
            matchMappingType = "string",
            mapping = @Field(type = FieldType.Keyword)
        ),
        @DynamicTemplate(
            name = "dates",
            matchMappingType = "date",
            mapping = @Field(type = FieldType.Date, format = "yyyy-MM-dd HH:mm:ss")
        )
    })

    // Getter、Setter省略
}

2. 创建Repository接口

import org.springframework.data.elasticsearch.repository.ElasticsearchRepository;

public interface MyIndexRepository extends ElasticsearchRepository<MyIndexEntity, String> {
}

3. 触发映射更新

在Spring Boot配置文件中开启映射自动同步（根据版本调整配置项）：

# Spring Boot 2.x
spring.data.elasticsearch.mapping.sync=true

或者手动通过ElasticsearchOperations触发更新：

import org.springframework.data.elasticsearch.core.ElasticsearchOperations;
import org.springframework.stereotype.Component;

import javax.annotation.PostConstruct;

@Component
public class MappingSyncHandler {
    private final ElasticsearchOperations elasticsearchOperations;

    public MappingSyncHandler(ElasticsearchOperations elasticsearchOperations) {
        this.elasticsearchOperations = elasticsearchOperations;
    }

    @PostConstruct
    public void updateMapping() {
        elasticsearchOperations.putMapping(MyIndexEntity.class);
    }
}

方式二：手动构建请求更新映射 #

如果不想依赖实体类，可直接通过RestHighLevelClient构建请求：

import org.elasticsearch.action.admin.indices.mapping.put.PutMappingRequest;
import org.elasticsearch.client.RequestOptions;
import org.elasticsearch.client.RestHighLevelClient;
import org.elasticsearch.common.xcontent.XContentBuilder;
import org.elasticsearch.common.xcontent.XContentFactory;
import org.springframework.stereotype.Component;

import javax.annotation.PostConstruct;
import java.io.IOException;

@Component
public class ManualMappingUpdater {
    private final RestHighLevelClient restHighLevelClient;

    public ManualMappingUpdater(RestHighLevelClient restHighLevelClient) {
        this.restHighLevelClient = restHighLevelClient;
    }

    @PostConstruct
    public void updateMapping() throws IOException {
        PutMappingRequest request = new PutMappingRequest("my-index");
        request.type("_doc");

        XContentBuilder builder = XContentFactory.jsonBuilder();
        builder.startObject()
                // 添加动态模板
                .startArray("dynamic_templates")
                    .startObject()
                        .startObject("strings")
                            .field("match_mapping_type", "string")
                            .startObject("mapping")
                                .field("type", "keyword")
                            .endObject()
                        .endObject()
                    .endObject()
                    .startObject()
                        .startObject("dates")
                            .field("match_mapping_type", "date")
                            .startObject("mapping")
                                .field("type", "date")
                                .field("format", "yyyy-MM-dd HH:mm:ss")
                            .endObject()
                        .endObject()
                    .endObject()
                .endArray()
                // 添加新增字段（原有字段无需重复定义）
                .startObject("properties")
                    .startObject("date").field("type", "long").endObject()
                    .startObject("favorite_color").field("type", "keyword").endObject()
                    .startObject("is_active").field("type", "boolean").endObject()
                    .startObject("number_float").field("type", "float").endObject()
                    .startObject("number_long").field("type", "long").endObject()
                    .startObject("updatedDate")
                        .field("type", "date")
                        .field("format", "yyyy-MM-dd HH:mm:ss")
                    .endObject()
                .endObject()
        .endObject();

        request.source(builder);
        restHighLevelClient.indices().putMapping(request, RequestOptions.DEFAULT);
    }
}

关键注意事项 #

1. 禁止修改已有字段类型：Elasticsearch不允许修改已有字段的映射类型（如integer改long），若需修改必须重新索引数据；
2. 动态模板生效范围：新增的动态模板仅对后续写入的文档生效，已有文档不会自动适配，如需适配需重新索引；
3. 版本兼容：确保Spring Data Elasticsearch版本与Elasticsearch服务端版本匹配，避免兼容性问题；
4. 增量更新特性：上述操作均为增量更新映射，不会重建索引，因此不会丢失现有数据。

内容的提问来源于stack exchange，提问作者spaceMan