使用 Laradock 安装 ElasticSearch
安装和使用
- 使用
docker-compose up
命令运行ElasticSearch
容器
docker-compose up -d elasticsearch
- 打开浏览器并通过端口
9200
访问本地主机 http://localhost:9200
默认用户是 user ,默认密码是 changeme
如果是在 laradock 中使用时
curl http://elasticsearch:9200
安装 ElasticSearch 插件
# 安装一个 ElasticSearch 插件
docker-compose exec elasticsearch /usr/share/elasticsearch/bin/elasticsearch-plugin install {plugin-name}
# 重启容器
docker-compose restart elasticsearch
安装 elasticsearch-analysis-ik 中文分词插件
比如,此时需要安装 elasticsearch-analysis-ik 中文分词插件,需要下载 ik 的 releases 源码 zip 包
# 方式1,你可以直接在 elasticsearch 容器外,执行以下命令
docker-compose exec elasticsearch /usr/share/elasticsearch/bin/elasticsearch-plugin install https://github.com/medcl/elasticsearch-analysis-ik/releases/download/v7.9.1/elasticsearch-analysis-ik-7.9.1.zip
# 方式2,你可以直接进入到 elasticsearch 容器内,然后执行以下命令
./bin/elasticsearch-plugin install https://github.com/medcl/elasticsearch-analysis-ik/releases/download/v7.9.1/elasticsearch-analysis-ik-7.9.1.zip
需要注意的是:如果你的 elasticsearch 的版本是 7.9.1 那么,你安装的 ik 插件也必须是 7.9.1 的版本,elasticsearch 的版本号可以通过访问 http://localhost:9200/ 查看 version.number
字段查看,然后 docker-compose restart elasticsearch
重启 elasticsearch 容器即可
安装 elasticsearch-analysis-ik 过程如下所示
[root@f1831cb3b4dd elasticsearch]# ./bin/elasticsearch-plugin install https://github.com/medcl/elasticsearch-analysis-ik/releases/download/v7.9.1/elasticsearch-analysis-ik-7.9.1.zip
-> Installing https://github.com/medcl/elasticsearch-analysis-ik/releases/download/v7.9.1/elasticsearch-analysis-ik-7.9.1.zip
-> Downloading https://github.com/medcl/elasticsearch-analysis-ik/releases/download/v7.9.1/elasticsearch-analysis-ik-7.9.1.zip
[=================================================] 100%??
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@ WARNING: plugin requires additional permissions @
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
* java.net.SocketPermission * connect,resolve
See http://docs.oracle.com/javase/8/docs/technotes/guides/security/permissions.html
for descriptions of what these permissions allow and the associated risks.
Continue with installation? [y/N]y
-> Installed analysis-ik
[root@f1831cb3b4dd elasticsearch]# ./bin/elasticsearch-plugin list
analysis-ik
查看插件列表
./bin/elasticsearch-plugin list
ElasticSearch 和 mysql 数据库的概念对比
MySQL | Elasticsearch |
---|---|
表(Table) | 索引(Index) |
记录(Row) | 文档(Document) |
字段(Column) | 字段(Fields) |
ElasticSearch 的简单使用
新建索引 index (创建表)
curl -XPUT http://localhost:9200/test_index
# 在 Elasticsearch 的返回中如果包含了 "acknowledged" : true, 则代表请求成功。
{"acknowledged":true,"shards_acknowledged":true,"index":"test_index"}
查看
curl http://localhost:9200/test_index
{"test_index":{"aliases":{},"mappings":{},"settings":{"index":{"creation_date":"1617069458624","number_of_shards":"1","number_of_replicas":"1","uuid":"XKjqatZTSOu9I_PiwzaNOQ","version":{"created":"7090199"},"provided_name":"test_index"}}}}
# 可以加上 pretty 参数,返回比较人性化的结构
curl http://localhost:9200/test_index\?pretty
{
"test_index" : {
"aliases" : { },
"mappings" : { },
"settings" : {
"index" : {
"creation_date" : "1617069458624",
"number_of_shards" : "1",
"number_of_replicas" : "1",
"uuid" : "XKjqatZTSOu9I_PiwzaNOQ",
"version" : {
"created" : "7090199"
},
"provided_name" : "test_index"
}
}
}
}
创建类型
对应的接口地址是 /{index_name}/_mapping
curl -H'Content-Type: application/json' -XPUT http://localhost:9200/test_index/_mapping?pretty -d'{
"properties": {
"title": { "type": "text", "analyzer": "ik_smart" },
"description": { "type": "text", "analyzer": "ik_smart" },
"price": { "type": "scaled_float", "scaling_factor": 100 }
}
}'
# 会返回
{
"acknowledged" : true
}
curl -H'Content-Type: application/json' -XPUT http://localhost:9200/products/_mapping/?pretty -d'{
"properties": {
"brand_id": { "type": "integer" },
"type": { "type": "integer" },
"title": { "type": "text", "analyzer": "ik_smart" },
"unit": { "type": "keyword" },
"sketch": { "type": "text", "analyzer": "ik_smart" },
"keywords": { "type": "text", "analyzer": "ik_smart" },
"tags": { "type": "keyword" },
"barcode": { "type": "keyword" },
"price": { "type": "scaled_float", "scaling_factor": 100 },
"market_price": { "type": "scaled_float", "scaling_factor": 100 },
"rating": { "type": "float" },
"sold_count": { "type": "integer" },
"review_count": { "type": "integer" },
"virtual_retail_num": { "type": "integer" },
"description": { "type": "text", "analyzer": "ik_smart" },
"stock": { "type": "integer" },
"warning_stock": { "type": "integer" },
"main_image": { "type": "keyword" },
"slider_image": { "type": "keyword" },
"status": { "type": "integer" },
"is_hot": { "type": "integer" },
"sort": { "type": "integer" },
"categories": {
"type": "nested",
"properties": {
"id": { "type": "integer", "copy_to": "categories_id" },
"pid": { "type": "integer" },
"name": { "type": "text", "analyzer": "ik_smart", "copy_to": "categories_name" },
"description": { "type": "text", "analyzer": "ik_smart", "copy_to": "categories_description" },
"status": { "type": "integer" },
"level": { "type": "integer" },
"img": { "type": "keyword" }
}
},
"brand": {
"type": "nested",
"properties": {
"id": { "type": "integer" },
"name": { "type": "text", "analyzer": "ik_smart", "copy_to": "brand_name" },
"description": { "type": "text", "analyzer": "ik_smart", "copy_to": "brand_description" },
"log_url": { "type": "keyword" },
"img": { "type": "keyword" }
}
},
"attrs": {
"type": "nested",
"properties": {
"id": { "type": "integer" },
"name": { "type": "keyword", "copy_to": "attrs_name" }
}
},
"skus": {
"type": "nested",
"properties": {
"id": { "type": "integer" },
"name": { "type": "text", "analyzer": "ik_smart"},
"main_url": { "type": "keyword" },
"price": { "type": "scaled_float", "scaling_factor": 100 },
"sold_count": { "type": "integer" }
}
}
}
}'
- 提交数据中的
properties
代表这个索引中各个字段的定义,其中 key 为字段名称,value 是字段的类型定义 type
定义了字段的数据类型,常用的有text
/integer
/date
/boolean
,还有更多类型keyword
,这是字符串类型的一种,这种类型是告诉 Elasticsearch 不需要对这个字段做分词,通常用于邮箱、标签、属性等字段。scaled_float
代表一个小数位固定的浮点型字段,与 Mysql 的 decimal 类型类似。scaling_factor
用来指定小数位精度,100 就代表精确到小数点后两位。nested
代表这个字段是一个复杂对象,由下一级的 properties 字段定义这个对象的字段。
analyzer
是一个新的概念,这是告诉 Elasticsearch 应该用什么方式去给这个字段做分词,这里我们用了ik_smart
,是一个中文分词器。copy_to
,Elasticsearch 的多字段匹配查询是不支持查询 Nested 对象的字段,但是我们又必须查询categories.name
字段,因此我们可以使用copy_to
参数,可以将categories.name
字段复制到上层,我们就可以通过categories_name
字段做多字段匹配查询
创建文档
对应的接口地址是 /{index_name}/_doc/{id} 这里的 id 和 mysql 中的 id 不一样,不是自增的,需要我们手动指定。
# 创建 id 为 1 的文档
curl -H'Content-Type: application/json' -XPUT http://localhost:9200/test_index/_doc/1?pretty -d'{
"title": "iPhone 7P",
"description": "iphone 第一批双摄像头",
"price": 6799
}'
# 会返回如下内容
{
"_index" : "test_index",
"_type" : "_doc",
"_id" : "1",
"_version" : 1,
"result" : "created",
"_shards" : {
"total" : 2,
"successful" : 1,
"failed" : 0
},
"_seq_no" : 0,
"_primary_term" : 2
}
# 创建 id 为 2 的文档
curl -H'Content-Type: application/json' -XPUT http://localhost:9200/test_index/_doc/2?pretty -d'{
"title": "OPPO find x",
"description": "高清像素",
"price": 3499
}'
# 会返回如下内容
{
"_index" : "test_index",
"_type" : "_doc",
"_id" : "2",
"_version" : 1,
"result" : "created",
"_shards" : {
"total" : 2,
"successful" : 1,
"failed" : 0
},
"_seq_no" : 1,
"_primary_term" : 2
}
读取文档数据
curl http://localhost:9200/test_index/_doc/1\?pretty
# 会返回如下内容
{
"_index" : "test_index",
"_type" : "_doc",
"_id" : "1",
"_version" : 1,
"_seq_no" : 0,
"_primary_term" : 2,
"found" : true,
"_source" : {
"title" : "iPhone 7P",
"description" : "iphone 第一批双摄像头",
"price" : 6799
}
}
查看 Elasticsearch 索引中有多少条数据
对应的接口地址为 /{index_name}/_doc/_count
curl http://localhost:9200/test_index/_doc/_count\?pretty
# 会返回如下内容
{
"count" : 3,
"_shards" : {
"total" : 1,
"successful" : 1,
"skipped" : 0,
"failed" : 0
}
}
简单搜索
curl -XPOST -H'Content-Type:application/json' http://localhost:9200/test_index/_doc/_search\?pretty -d'
{
"query" : { "match" : { "description" : "iphone" }}
}'
# 会返回如下内容
{
"took" : 16,
"timed_out" : false,
"_shards" : {
"total" : 1,
"successful" : 1,
"skipped" : 0,
"failed" : 0
},
"hits" : {
"total" : {
"value" : 1,
"relation" : "eq"
},
"max_score" : 0.60996956,
"hits" : [
{
"_index" : "test_index",
"_type" : "_doc",
"_id" : "1",
"_score" : 0.60996956,
"_source" : {
"title" : "iPhone 7P",
"description" : "iphone 第一批双摄像头",
"price" : 6799
}
}
]
}
}
原文链接地址
**粗体** _斜体_ [链接](http://example.com) `代码` - 列表 > 引用
。你还可以使用@
来通知其他用户。