Question

我正在尝试fake index per user，但我的用户ID是电子邮件。当我的别名中有@时，搜索为空。

例如，如果我创建一个索引my_index和两个别名my_alias和my@lias，如下所示，my_alias搜索工作正常，我的@ lias不是：

curl -XPUT 'elasticsearch:9200/my_index'
{"acknowledged":true}

curl -XPUT 'elasticsearch:9200/my_index/_alias/my_alias -d '{"routing": "my_alias", "filter": {"term": {"doc_id": "my_alias"}}}'
{"acknowledged":true}

curl -XPUT 'elasticsearch:9200/my_index/_alias/my@lias' -d '{"routing": "my@lias", "filter": {"term": {"doc_id": "my@lias"}}}'
{"acknowledged":true}

curl -XPUT 'elasticsearch:9200/my_alias/doc/1' -d '{doc_id:"my_alias", title: "hello world"}'
{"_index":"my_index","_type":"doc","_id":"1","_version":1,"_shards":{"total":2,"successful":1,"failed":0},"created":true}

curl -XPUT 'elasticsearch:9200/my@lias/doc/2' -d '{doc_id:"my@lias", title: "hello dude"}'
{"_index":"my_index","_type":"doc","_id":"2","_version":1,"_shards":{"total":2,"successful":1,"failed":0},"created":true}

curl -XGET 'elasticsearch:9200/my_alias/doc/_search?pretty=true' -d'{"query": {"match": {"title": "hello"}}}'
{
    "took" : 1,
    "timed_out" : false,
    "_shards" : {
        "total" : 1,
        "successful" : 1,
        "failed" : 0
    },
    "hits" : {
        "total" : 1,
        "max_score" : 0.19178301,
        "hits" : [ {
            "_index" : "my_index",
            "_type" : "doc",
            "_id" : "1",
            "_score" : 0.19178301,
            "_routing" : "my_alias",
            "_source" : {
                "doc_id" : "my_alias",
                "title" : "hello world"
            }
        } ]
    }
}

curl -XGET 'elasticsearch:9200/my@lias/doc/_search?pretty=true' -d'{"query": {"match": {"title": "hello"}}}'
{
    "took" : 1,
    "timed_out" : false,
    "_shards" : {
        "total" : 1,
        "successful" : 1,
        "failed" : 0
    },
    "hits" : {
        "total" : 0,
        "max_score" : null,
        "hits" : [ ]
    }
}

让我感到困惑的是，没有Invalid Alias Name Exception，搜索结果还可以，但是空的。

我还尝试使用url-encoding %40：elasticsearch:9200/my%40lias，但它是一样的。

如果我搜索整个索引，我会看到这两个文件。如果我使用@创建索引，索引搜索也可以。如果我卷曲_aliases以查看ES状态，我可以看到它：

"my_index" : {
  "aliases" : {
    "my@lias" : {
      "filter" : {
        "term" : {
          "doc_id" : "my@lias"
        }
      },
      "index_routing" : "my@lias",
      "search_routing" : "my@lias"
    },
    "my_alias" : {
      "filter" : {
        "term" : {
          "doc_id" : "my_alias"
        }
      },
      "index_routing" : "my_alias",
      "search_routing" : "my_alias"
    }
  }
},

你认为这种行为有什么理由吗？我想念一下吗？

Answer 1

您只需对@符号进行网址编码，如下所示：

curl -XGET 'elasticsearch:9200/my%40lias/doc/_search?pretty=true'
                                  ^
                                  |
                             change this

Answer 2

我在本地复制了您的索引，我认为您遇到的问题是您没有正确索引文档。

例如，看一下这个请求：

curl -XPUT 'elasticsearch:9200/my_alias/doc/1' -d '{doc_id:"my_alias", title: "hello world"}'

应该是

curl -XPUT "http://localhost:9200/my_alias/doc/1" -d'
{"doc_id":"my_alias", "title": "hello world"}'

请注意 {doc_id：＆＃34; my_alias＆＃34; 和 {＆＃34; doc_id＆＃34;：＆＃34; my_alias＆＃34; < / strong>因为您缺少doc_id的引号。这也适用于标题。

完全固定的例子：

PUT /my_index PUT /my_index/_alias/my_alias {"routing": "my_alias", "filter": {"term": {"doc_id": "my_alias"}}} PUT /my_index/_alias/my@lias {"routing": "my@lias", "filter": {"term": {"doc_id": "my@lias"}}} PUT /my_alias/doc/1 { "doc_id":"my_alias", "title": "hello world" } PUT /my@lias/doc/2 { "doc_id": "my@lias", "title": "hello dude" } GET /my_alias/doc/_search?pretty=true { "query": { "match": { "title": "hello" } } } { "took": 2, "timed_out": false, "_shards": { "total": 1, "successful": 1, "failed": 0 }, "hits": { "total": 1, "max_score": 0.19178301, "hits": [ { "_index": "my_index", "_type": "doc", "_id": "1", "_score": 0.19178301, "_routing": "my_alias", "_source": { "doc_id": "my_alias", "title": "hello world" } } ] } }

弹性搜索别名是否有禁用字符？

2 个答案: