搜索引擎之Elastic Search的高级使用
在开发中，随着业务需求的迭代，较⽼的业务逻辑就要⾯临更新甚⾄是重构，⽽对于es来说，为了适应新的业务逻辑，可能就要对原有的索引做⼀些修改，⽐如对某些字段做调整，甚⾄是重建索引。⽽做这些操作的时候，可能会对业务造成影响，甚⾄是停机调整等问题。由此，es提供了索引别名来解决这些问题。 索引别名就像⼀个快捷⽅式或是软连接，可以指向⼀个或多个索引，也可以给任意⼀个需要索引名的API来使⽤。别名的应⽤为程序提供了极⼤地灵活性。
查询别名GET /nba/_alias
GET /_alias新增别名POST /_aliases
{
"actions": [
{
"add": {
"index": "nba",
"alias": "nba_v1.0"
}
}
]
}PUT /nba/_alias/nba_v1.1删除别名POST /_aliases
{
"actions": [
{
"remove": {
"index": "nba",
"alias": "nba_v1.0"
}
}
]
}DELETE /nba/_alias/nba_v1.1重命名POST /_aliases
{
"actions": [
{
"remove": {
"index": "nba",
"alias": "nba_v1.0"
}
},
{
"add": {
"index": "nba",
"alias": "nba_v2.0"
}
}
]
}为多个索引指定⼀个别名POST /_aliases
{
"actions": [
{
"add": {
"index": "nba",
"alias": "national_player"
}
},
{
"add": {
"index": "wnba",
"alias": "national_player"
}
}
]
}为同个索引指定多个别名POST /_aliases
{
"actions": [
{
"add": {
"index": "nba",
"alias": "nba_v2.1"
}
},
{
"add": {
"index": "nba",
"alias": "nba_v2.2"
}
}
]
}通过别名读索引当别名指定了⼀个索引，则查出⼀个索引。
GET /nba_v2.1当别名指定了多个索引，则查出多个索引。
GET /national_player通过别名写索引当别名指定了⼀个索引，则可以做写的操作。
POST /nba_v2.1/_doc/566
{
"countryEn": "Croatia",
"teamName": "快船",
"birthDay": 858661200000,
"country": "克罗地亚",
"teamCityEn": "LA",
"code": "ivica_zubac",
"displayAffiliation": "Croatia",
"displayName": "伊维察 祖巴茨哥哥",
"schoolType": "",
"teamConference": "⻄部",
"teamConferenceEn": "Western",
"weight": "108.9 公⽄",
"teamCity": "洛杉矶",
"playYear": 3,
"jerseyNo": "40",
"teamNameEn": "Clippers",
"draft": 2016,
"displayNameEn": "Ivica Zubac",
"heightValue": 2.16,
"birthDayStr": "1997-03-18",
"position": "中锋",
"age": 22,
"playerId": "1627826"
}当别名指定了多个索引，可以指定写某个索引。
POST /_aliases
{
"actions": [
{
"add": {
"index": "nba",
"alias": "national_player",
"is_write_index": true
}
},
{
"add": {
"index": "wnba",
"alias": "national_player"
}
}
]
}POST /national_player/_doc/566
{
"countryEn": "Croatia",
"teamName": "快船",
"birthDay": 858661200000,
"country": "克罗地亚",
"teamCityEn": "LA",
"code": "ivica_zubac",
"displayAffiliation": "Croatia",
"displayName": "伊维察 祖巴茨妹妹",
"schoolType": "",
"teamConference": "⻄部",
"teamConferenceEn": "Western",
"weight": "108.9 公⽄",
"teamCity": "洛杉矶",
"playYear": 3,
"jerseyNo": "40",
"teamNameEn": "Clippers",
"draft": 2016,
"displayNameEn": "Ivica Zubac",
"heightValue": 2.16,
"birthDayStr": "1997-03-18",
"position": "中锋",
"age": 22,
"playerId": "1627826"
}es之如何重建索引
背景Elasticsearch是⼀个实时的分布式搜索引擎，为⽤户提供搜索服务，当我们决定存储某种数据时，在创建索引的时候需要将数据结构完整确定下来，于此同时索引的设定和很多固定配置将⽤不能改变。当需要改变数据结构时，就需要重新建⽴索引，为此，Elastic团队提供了很多辅助⼯具帮助开发⼈员进⾏重建索引。
步骤nba取⼀个别名nba_latest, nba_latest作为对外使⽤。新增⼀个索引nba_20220101，结构复制于nba索引，根据业务要求修改字段。将nba数据同步到nba_20220101。给nba_20220101添加别名nba_latest，删除nba别名nba_latest。删除nba索引。我们对外提供访问nba索引时使⽤的是nba_latest别名1.新增⼀个索引(⽐如修改字段类型，jerseyNo改成keyword类型)
PUT /nba_20220101
{
"mappings": {
"properties": {
"age": {
"type": "integer"
},
"birthDay": {
"type": "date"
},
"birthDayStr": {
"type": "keyword"
},
"code": {
"type": "text"
},
"country": {
"type": "keyword"
},
"countryEn": {
"type": "keyword"
},
"displayAffiliation": {
"type": "text"
},
"displayName": {
"type": "text"
},
"displayNameEn": {
"type": "text"
},
"draft": {
"type": "long"
},
"heightValue": {
"type": "float"
},
"jerseyNo": {
"type": "keyword"
},
"playYear": {
"type": "long"
},
"playerId": {
"type": "keyword"
},
"position": {
"type": "text"
},
"schoolType": {
"type": "text"
},
"teamCity": {
"type": "text"
},
"teamCityEn": {
"type": "text"
},
"teamConference": {
"type": "keyword"
},
"teamConferenceEn": {
"type": "keyword"
},
"teamName": {
"type": "keyword"
},
"teamNameEn": {
"type": "keyword"
},
"weight": {
"type": "text"
}
}
}
}2.将旧索引数据copy到新索引
同步等待，接⼝将会在 reindex 结束后返回。
POST /_reindex
{
"source": {
"index": "nba"
},
"dest": {
"index": "nba_20220101"
}
}异步执⾏，如果 reindex 时间过⻓，建议加上 wait_for_completion=false 的参数条件，这样 reindex 将直接返回 taskId
POST /_reindex?wait_for_completion=false
{
"source": {
"index": "nba"
},
"dest": {
"index": "nba_20220101"
}
}3.替换别名
POST /_aliases
{
"actions": [
{
"add": {
"index": "nba_20220101",
"alias": "nba_latest"
}
},
{
"remove": {
"index": "nba",
"alias": "nba_latest"
}
}
]
}4.删除旧索引
DELETE /nba5.通过别名访问新索引
POST /nba_latest/_search
{
"query": {
"match": {
"displayNameEn": "james"
}
}
}es之refresh操作理想的搜索：
新的数据⼀添加到索引中⽴⻢就能搜索到，但是真实情况不是这样的。我们使⽤链式命令请求，先添加⼀个⽂档，再⽴刻搜索。
curl -X PUT localhost:9200/star/_doc/888 -H Content-Type:
application/json -d { "displayName": "蔡徐坤" }
curl -X GET localhost:9200/star/_doc/_search?pretty强制刷新
curl -X PUT localhost:9200/star/_doc/666?refresh -H Content-Type:
application/json -d { "displayName": "杨超越" }
curl -X GET localhost:9200/star/_doc/_search?pretty修改默认更新时间(默认时间是1s)
PUT /star/_settings
{
"index": {
"refresh_interval": "5s"
}
}将refresh关闭
PUT /star/_settings
{
"index": {
"refresh_interval": "-1"
}
}es之高亮查询前⾔
如果返回的结果集中很多符合条件的结果，那怎么能⼀眼就能看到我们想要的那个结果呢？⽐如下⾯⽹站所示的那样，我们搜索 ⼩d课堂 ，在结果集中，将所有 ⼩d课堂 ⾼亮显示？
高亮查询
POST /nba_latest/_search
{
"query": {
"match": {
"displayNameEn": "james"
}
},
"highlight": {
"fields": {
"displayNameEn": {}
}
}
}自定义高亮查询
POST /nba_latest/_search
{
"query": {
"match": {
"displayNameEn": "james"
}
},
"highlight": {
"fields": {
"displayNameEn": {
"pre_tags": [
"
"
],
"post_tags": [
""
]
}
}
}
}es之查询建议查询建议是什么
查询建议，是为了给⽤户提供更好的搜索体验。包括：词条检查，⾃动补全。词条检查
Suggester
Term suggesterPhrase suggesterCompletion suggester字段
Term suggester
term 词条建议器，对给输⼊的⽂本进⾏分词，为每个分词提供词项建议。
POST /nba_latest/_search
{
"suggest": {
"my-suggestion": {
"text": "jamse hardne",
"term": {
"suggest_mode": "missing",
"field": "displayNameEn"
}
}
}
}Phrase suggester
phrase 短语建议，在term的基础上，会考量多个term之间的关系，⽐如是否同时出现在索引的原⽂⾥，相邻程度，以及词频等。
POST /nba_latest/_search
{
"suggest": {
"my-suggestion": {
"text": "jamse harden",
"phrase": {
"field": "displayNameEn"
}
}
}
}Completion suggester
Completion 完成建议。
POST /nba_latest/_search
{
"suggest": {
"my-suggestion": {
"text": "Miam",
"completion": {
"field": "teamCityEn"
}
}
}
}扫一扫，关注我们
