Elasticsearch、Kibana 使用

通过最近一段时间对`ES`的基础使用，也算有了一定的了解，特在此记录一下，避免以后忘记。

ES、Kibana、Elasticsearch-head安装

docker一键部署

docker-compose.yaml文件

version: '3'

services:
  es:
    image: elasticsearch:7.6.2
    restart: always
    ports:
      - "9200:9200"
      - "9300:9300"
    environment:
      - "discovery.type=single-node"
      - "ES_JAVA_OPTS=-Xms64m -Xmx512m"
    volumes:
      - esdata:/usr/share/elasticsearch/data
      - esplugins:/usr/share/elasticsearch/plugins
    user: root
    networks:
      - es
  kibana:
    image: kibana:7.6.2
    restart: always
    environment:
      - ELASTICSEARCH_HOSTS=http://elasticsearch:9200
      - I18N_LOCALE=zh-CN
#      - ELASTICSEARCH_USERNAME=elastic
#      - ELASTICSEARCH_PASSWORD=mypasword
    depends_on:
      - es
    ports:
      - "5601:5601"
    links:
      - "es:elasticsearch"
    networks:
      - es

volumes:
  esdata:
    driver: local
  esplugins:
    driver: local
networks:
  es:

快速使用

启动：docker-compose up -d

删除：docker-compose down

查看容器信息：docker inspect container-name-xxx

安装ik分词

docker volume ls 查看所有挂载

docker volume inspect es_esplugins查看挂载信息后，找到挂载地址Mountpoint对应地址。

root@dns:~/downloads/es_search_fastapi# docker volume inspect es_esplugins
[
    {
        "CreatedAt": "2021-12-22T15:42:14+08:00",
        "Driver": "local",
        "Labels": {
            "com.docker.compose.project": "es",
            "com.docker.compose.version": "1.29.2",
            "com.docker.compose.volume": "esplugins"
        },
        "Mountpoint": "/var/lib/docker/volumes/es_esplugins/_data",
        "Name": "es_esplugins",
        "Options": null,
        "Scope": "local"
    }
]

1. 在 https://github.com/medcl/elasticsearch-analysis-ik 下载对应版本
2. 解压到 /var/lib/docker/volumes/es_esplugins/_data/ik目录下(没有ik目录，就自己创建一个)
3. 重启 ES容器 docker restart es_es_1，就可以了

Elasticsearch-head 插件

我使用的是[谷歌插件](https://chrome.google.com/webstore/detail/elasticsearch-head/ffmkiejjmecolpfloofpjologoblkegm?utm_source=chrome-ntp-icon),可以自行搜索下载。

Kibana对ES操作使用

操作

本来想写的，但是觉得这样内容太冗余了。后面放一些我收藏的别人的写的吧，毕竟一直重复也没有任何意义。

注意事项

1. ik分词会对内容分词，因此需要全匹配的不适用ik和keyword
2. _reindex时，内容过多会超时，因此加上?wait_for_completion=false，不用等待
3. ES分页修改查询10000限制，index.max_result_window 下面有修改语句
4. ES7.0版本以后，一个index对应一个type了，不再是以前的一个index对应多个type，因此index也相当于一张表了

代码记录

GET _search
{
  "query": {
    "match_all": {}
  }
}
# 修改查询10000条限制
PUT data_item/_settings
{
  "index.max_result_window":1000000
}
# 获取最后一个id，保持增量更新
GET data_item/_search
{
  "_source": "item_id", 
  "sort": [
    {
      "item_id": {
        "order": "desc"
      }
    }
  ]
  , "size": 1
}
# 搜索推荐词 推荐
GET data_item/_search
{ 
  "suggest": {
    "my-suggestion": {
      "text": "科技",
      "term": {
        "suggest_mode": "popular",
        "field": "item_title"
      }
    }
  }
}
# 获取索引映射
GET data_item/_mapping
# 删除索引
DELETE data_item
# 新建索引
PUT data_item
{
    "mappings" : {
      "properties" : {
        "area_code" : {
          "type" : "integer"
        },
        "create_time" : {
          "type" : "date"
        },
        "dept_id" : {
          "type" : "long"
        },
        "dept_name" : {
          "type" : "text",
          "analyzer" : "ik_max_word",
          "search_analyzer" : "ik_smart"
        },
        "item_content" : {
          "type" : "text",
          "analyzer" : "ik_max_word",
          "search_analyzer" : "ik_smart"
        },
        "item_deadline" : {
          "type" : "date"
        },
        "item_id" : {
          "type" : "long"
        },
        "item_id_key" : {
          "type" : "text",
          "index" : false
        },
        "item_imgurl" : {
          "type" : "text",
          "index" : false
        },
        "item_pulishdate" : {
          "type" : "date"
        },
        "item_title" : {
          "type" : "text",
          "analyzer" : "ik_max_word",
          "search_analyzer" : "ik_smart"
        },
        "item_type" : {
          "type" : "long"
        },
        "item_type_name" : {
          "type" : "text"
        },
        "item_url" : {
          "type" : "text",
          "index" : false
        },
        "level_code" : {
          "type" : "long"
        },
        "level_code_name" : {
          "type" : "text"
        }
      }
    }
}
# 索引内容转移到另一个索引
POST _reindex?wait_for_completion=false
{
  "source": {
    "index": "data_item",
    "size": 5000
  },
  "dest": {
    "index": "data_item_project"
  }
}

# 目前我这里查询最复杂的一种
GET data_item/_search
{
  "query": {
    "bool": {
      "must": [
        {
          "term": {
            "dept_id": 8
          }
        },
        {
          "term": {
            "item_type": 1
          }
        },
        {
          "term": {
            "level_code": 2
          }
        },
        {
          "match": {
            "item_title": "创新"
          }
        }
      ],
      "filter": {
        "range": {
          "item_pulishdate": {
            "gte": "2021-01-11"
          }
        }
      },
      "should": [
        {
          "term": {
            "area_code": "510100"
          }
        },
        {
          "term": {
            "area_code": "510300"
          }
        }
      ]
    }
  },
  "highlight": {
    "pre_tags": [
      "<em>"
    ],
    "post_tags": [
      "</em>",
    ],
    "fields": {
      "item_title": {}
    }
  },
  "from": 0,
  "size": 10
}