【发布时间】：2015-07-26 05:51:32
【问题描述】：
我对弹性搜索很陌生，如果这是一个微不足道的问题，我深表歉意。
我有一个时间序列，每 n 秒不定期更新一次，我想在历史上绘制。数据包含一个称为“分数”的长变量，以及一个长变量，其中每个“分数”的时间点称为“时间”作为时间戳。
为了减少长时间尺度图中的点数（例如一整年），我想汇总 256 个桶中的数据，并为每个桶使用最大“分数”值；但是，我需要保留每个分数的原始时间戳，而不是存储桶的开头。
我设法通过运行以下查询来获取存储桶：
curl -XGET 'http://localhost:9200/localhost.localdomain/SET_APPS/_search' -d'
{
"query" : {
"range" : {
"time" : {
"from" : 1429010378445,
"to" : 1431602378445,
"include_lower" : true,
"include_upper" : true
}
}
},
"aggregations" : {
"time_hist" : {
"histogram" : {
"field" : "time",
"interval" : 10125000,
"order" : {
"_count" : "asc"
},
"min_doc_count" : 0,
"extended_bounds" : {
"min" : 1429010378445,
"max" : 1431602378445
}
},
"aggregations" : {
"max_score" : {
"max" : {
"field" : "score"
}
}
}
}
}
}
}'但是，我只获取存储桶的时间戳，而我需要分数的原始时间：
{
"took": 8,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 4,
"failed": 1,
"failures": [{
"index": "localhost.localdomain",
"shard": 2,
"status": 500,
"reason
": "QueryPhaseExecutionException[[localhost.localdomain][2]: query[filtered(time:[1429010378445 TO 1431602378445])->cache(_type:SET_APPS)],from[0],size
[10]: Query Failed [Failed to execute main query]]; nested: IllegalStateException[unexpected docvalues type NONE for field 'score' (expected one of [S
ORTED_NUMERIC, NUMERIC]). Use UninvertingReader or index with docvalues.]; "
}]
},
"hits": {
"total": 2018,
"max_score": 1.0,
"hits": [{
"_index": "localhost.localdomain",
"_type": "SET_APPS",
"_id": "AU09dUBR80Hb_Fungv_r",
"_score": 1.0,
"_source": {
time: 1431255203918,
score: 6027
}
}, {
"_index": "localhost.localdomain",
"_type": "SET_APPS",
"_id": "AU09c7MS80Hb_Fungv_X",
"_score": 1.0,
"_source": {
time: 1431255102221,
score: 5518,
}
}
....
]
},
"aggregations": {
"time_hist": {
"buckets": [{
"key": 1429002000000,
"doc_count": 0,
"max_score": {
"value": null
}
},
......
{
"key": 1431249750000,
"doc_count": 215,
"max_score": {
"value": 8564.0,
"value_as_string": "8564.0"
}
}, {
"key": 1431280125000,
"doc_count": 228,
"max_score": {
"value": 18602.0,
"value_as_string": "18602.0"
}
}, {
"key": 1431259875000,
"doc_count": 658,
"max_score": {
"value": 17996.0,
"value_as_string": "17996.0"
}
}, {
"key": 1431270000000,
"doc_count": 917,
"max_score": {
"value": 17995.0,
"value_as_string": "17995.0"
}
}]
}
}
}在上面的结果中，如果我们专门查询分数 18602，我们得到的是真正的时间戳：
$ curl -XGET 'http://localhost:9200/localhost.localdomain/SET_APPS/_search' -d'
{
"fields": [ "time", "score" ],
"query" : {
"term": {
"score": "18602"
}
}
}'
{"took":3,"timed_out":false,"_shards":{"total":5,"successful":5,"failed":0},"hits":{"total":1,"max_score":1.0,"hits":[{"_index":"localhost.localdomain
","_type":"SET_APPS","_id":"AU0-90Vsi-vs_2ajcYu-","_score":1.0,"fields":{"score":[18602],"time":[1431280502124]}}]}}感谢任何帮助！
【问题讨论】：
标签： java elasticsearch max histogram aggregation