我的查询有什么问题吗.Nest elastic C#
Is Anything wrong in my query .Nest elastic C#
我的 .Nest 库查询有问题吗?我的查询将获取所有数据,我需要按多个术语获取。
我想要的查询字符串弹性结果:
{
"took": 2,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"failed": 0
},
"hits": {
"total": 1000,
"max_score": 0,
"hits": []
},
"aggregations": {
"log_query": {
"doc_count": 2,
"histogram_Log": {
"buckets": [
{
"key_as_string": "06/02/2015 12:00:00",
"key": 1423180800000,
"doc_count": 1
},
{
"key_as_string": "21/02/2015 12:00:00",
"key": 1424476800000,
"doc_count": 1
}
]
}
}
}
}
我的查询字符串弹性:
{
"size": 0,
"aggs": {
"log_query": {
"filter": {
"bool": {
"must": [
{
"term": {
"cluster": "giauht1"
}
},
{
"term": {
"server": "hadoop0"
}
},
{
"term": {
"type": "Warn"
}
},
{
"range": {
"actionTime": {
"gte": "2015-02-01",
"lte": "2015-02-24"
}
}
}
]
}
},
"aggs": {
"histogram_Log": {
"date_histogram": {
"field": "actionTime",
"interval": "1d",
"format": "dd/MM/YYYY hh:mm:ss"
}
}
}
}
}
}
我的 .nest 库查询:
Func<SearchDescriptor<LogInfoIndexView>, SearchDescriptor<LogInfoIndexView>> query =
que => que.Aggregations(aggs => aggs.Filter("log_query", fil =>
{
fil.Filter(fb => fb.Bool(fm => fm.Must(
ftm =>
{
ftm.Term(t => t.Cluster, cluster);
ftm.Term(t => t.Server, server);
ftm.Term(t => t.Type, logLevel);
ftm.Range(r => r.OnField("actionTime").GreaterOrEquals(from.Value).LowerOrEquals(to.Value));
return ftm;
}))).Aggregations(faggs => faggs.DateHistogram("histogram_Log", dr =>
{
dr.Field("actionTime");
dr.Interval("1d");
dr.Format("dd/MM/YYYY hh:mm:ss");
return dr;
}));
return fil;
})).Size(0).Type(new LogInfoIndexView().TypeName);
var result = client.Search(query);
我的.nest 结果:
我的模型映射:
{
"onef-sora": {
"mappings": {
"FPT.OneF.Api.Log": {
"properties": {
"actionTime": {
"type": "date",
"format": "dateOptionalTime"
},
"application": {
"type": "string",
"index": "not_analyzed"
},
"cluster": {
"type": "string",
"index": "not_analyzed"
},
"detail": {
"type": "string",
"index": "not_analyzed"
},
"iD": {
"type": "string"
},
"message": {
"type": "string",
"index": "not_analyzed"
},
"server": {
"type": "string",
"index": "not_analyzed"
},
"source": {
"type": "string",
"index": "not_analyzed"
},
"tags": {
"type": "string",
"index": "not_analyzed"
},
"type": {
"type": "string",
"index": "not_analyzed"
},
"typeLog": {
"type": "string"
},
"typeName": {
"type": "string"
},
"url": {
"type": "string",
"index": "not_analyzed"
},
"user": {
"type": "string",
"index": "not_analyzed"
}
}
}
}
}
}
传递给 Bool()
过滤器的 Must()
条件采用 params Func<FilterDescriptor<T>, FilterContainer>[]
,但在您的过滤器中,Term()
和 Range()
过滤器链接到相同的过滤器实例;不幸的是,这并不像您 可能 期望的那样工作,最终结果实际上是一个空的 json 对象传递给查询 DSL 中的 must
子句过滤,即你最终得到
{
"size": 0,
"aggs": {
"log_query": {
"filter": {
"bool": {
"must": [
{} /* where are the filters?! */
]
}
},
"aggs": {
"histogram_Log": {
"date_histogram": {
"field": "actionTime",
"interval": "1d",
"format": "dd/MM/YYYY hh:mm:ss"
}
}
}
}
}
}
解决方法是传递一个Func<FilterDescriptor<T>, FilterContainer>
的数组;以下符合您的查询 DSL
void Main()
{
var settings = new ConnectionSettings(new Uri("http://localhost:9200"));
var connection = new InMemoryConnection(settings);
var client = new ElasticClient(connection: connection);
DateTime? from = new DateTime(2015, 2,1);
DateTime? to = new DateTime(2015, 2, 24);
var docs = client.Search<LogInfoIndexView>(s => s
.Size(0)
.Type("type")
.Aggregations(a => a
.Filter("log_query", f => f
.Filter(ff => ff
.Bool(b => b
.Must(m => m
.Term(t => t.Cluster, "giauht1"),
m => m
.Term(t => t.Server, "hadoop0"),
m => m
.Term(t => t.Type, "Warn"),
m => m
.Range(r => r.OnField("actionTime").GreaterOrEquals(from.Value).LowerOrEquals(to.Value))
)
)
)
.Aggregations(aa => aa
.DateHistogram("histogram_Log", da => da
.Field("actionTime")
.Interval("1d")
.Format("dd/MM/YYYY hh:mm:ss")
)
)
)
)
);
Console.WriteLine(Encoding.UTF8.GetString(docs.RequestInformation.Request));
}
public class LogInfoIndexView
{
public string Cluster { get; set; }
public string Server { get; set; }
public string Type { get; set; }
public DateTime ActionTime { get; set; }
}
返回
{
"size": 0,
"aggs": {
"log_query": {
"filter": {
"bool": {
"must": [
{
"term": {
"cluster": "giauht1"
}
},
{
"term": {
"server": "hadoop0"
}
},
{
"term": {
"type": "Warn"
}
},
{
"range": {
"actionTime": {
"lte": "2015-02-24T00:00:00.000",
"gte": "2015-02-01T00:00:00.000"
}
}
}
]
}
},
"aggs": {
"histogram_Log": {
"date_histogram": {
"field": "actionTime",
"interval": "1d",
"format": "dd/MM/YYYY hh:mm:ss"
}
}
}
}
}
}
编辑:
在回答您的评论时,filtered query filter
and a filter aggregation
之间的区别在于前者在查询阶段开始时将过滤应用于所有文档,并且过滤器通常被缓存,从而提高后续查询的性能过滤器,而后者适用于聚合范围,将当前上下文中的文档过滤到单个桶中。如果您的查询只是为了执行聚合并且您可能 运行 使用相同的过滤器进行聚合,我认为 filtered query filter
应该提供更好的性能。
我的 .Nest 库查询有问题吗?我的查询将获取所有数据,我需要按多个术语获取。 我想要的查询字符串弹性结果:
{
"took": 2,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"failed": 0
},
"hits": {
"total": 1000,
"max_score": 0,
"hits": []
},
"aggregations": {
"log_query": {
"doc_count": 2,
"histogram_Log": {
"buckets": [
{
"key_as_string": "06/02/2015 12:00:00",
"key": 1423180800000,
"doc_count": 1
},
{
"key_as_string": "21/02/2015 12:00:00",
"key": 1424476800000,
"doc_count": 1
}
]
}
}
}
}
我的查询字符串弹性:
{
"size": 0,
"aggs": {
"log_query": {
"filter": {
"bool": {
"must": [
{
"term": {
"cluster": "giauht1"
}
},
{
"term": {
"server": "hadoop0"
}
},
{
"term": {
"type": "Warn"
}
},
{
"range": {
"actionTime": {
"gte": "2015-02-01",
"lte": "2015-02-24"
}
}
}
]
}
},
"aggs": {
"histogram_Log": {
"date_histogram": {
"field": "actionTime",
"interval": "1d",
"format": "dd/MM/YYYY hh:mm:ss"
}
}
}
}
}
}
我的 .nest 库查询:
Func<SearchDescriptor<LogInfoIndexView>, SearchDescriptor<LogInfoIndexView>> query =
que => que.Aggregations(aggs => aggs.Filter("log_query", fil =>
{
fil.Filter(fb => fb.Bool(fm => fm.Must(
ftm =>
{
ftm.Term(t => t.Cluster, cluster);
ftm.Term(t => t.Server, server);
ftm.Term(t => t.Type, logLevel);
ftm.Range(r => r.OnField("actionTime").GreaterOrEquals(from.Value).LowerOrEquals(to.Value));
return ftm;
}))).Aggregations(faggs => faggs.DateHistogram("histogram_Log", dr =>
{
dr.Field("actionTime");
dr.Interval("1d");
dr.Format("dd/MM/YYYY hh:mm:ss");
return dr;
}));
return fil;
})).Size(0).Type(new LogInfoIndexView().TypeName);
var result = client.Search(query);
我的.nest 结果:
我的模型映射:
{
"onef-sora": {
"mappings": {
"FPT.OneF.Api.Log": {
"properties": {
"actionTime": {
"type": "date",
"format": "dateOptionalTime"
},
"application": {
"type": "string",
"index": "not_analyzed"
},
"cluster": {
"type": "string",
"index": "not_analyzed"
},
"detail": {
"type": "string",
"index": "not_analyzed"
},
"iD": {
"type": "string"
},
"message": {
"type": "string",
"index": "not_analyzed"
},
"server": {
"type": "string",
"index": "not_analyzed"
},
"source": {
"type": "string",
"index": "not_analyzed"
},
"tags": {
"type": "string",
"index": "not_analyzed"
},
"type": {
"type": "string",
"index": "not_analyzed"
},
"typeLog": {
"type": "string"
},
"typeName": {
"type": "string"
},
"url": {
"type": "string",
"index": "not_analyzed"
},
"user": {
"type": "string",
"index": "not_analyzed"
}
}
}
}
}
}
传递给 Bool()
过滤器的 Must()
条件采用 params Func<FilterDescriptor<T>, FilterContainer>[]
,但在您的过滤器中,Term()
和 Range()
过滤器链接到相同的过滤器实例;不幸的是,这并不像您 可能 期望的那样工作,最终结果实际上是一个空的 json 对象传递给查询 DSL 中的 must
子句过滤,即你最终得到
{
"size": 0,
"aggs": {
"log_query": {
"filter": {
"bool": {
"must": [
{} /* where are the filters?! */
]
}
},
"aggs": {
"histogram_Log": {
"date_histogram": {
"field": "actionTime",
"interval": "1d",
"format": "dd/MM/YYYY hh:mm:ss"
}
}
}
}
}
}
解决方法是传递一个Func<FilterDescriptor<T>, FilterContainer>
的数组;以下符合您的查询 DSL
void Main()
{
var settings = new ConnectionSettings(new Uri("http://localhost:9200"));
var connection = new InMemoryConnection(settings);
var client = new ElasticClient(connection: connection);
DateTime? from = new DateTime(2015, 2,1);
DateTime? to = new DateTime(2015, 2, 24);
var docs = client.Search<LogInfoIndexView>(s => s
.Size(0)
.Type("type")
.Aggregations(a => a
.Filter("log_query", f => f
.Filter(ff => ff
.Bool(b => b
.Must(m => m
.Term(t => t.Cluster, "giauht1"),
m => m
.Term(t => t.Server, "hadoop0"),
m => m
.Term(t => t.Type, "Warn"),
m => m
.Range(r => r.OnField("actionTime").GreaterOrEquals(from.Value).LowerOrEquals(to.Value))
)
)
)
.Aggregations(aa => aa
.DateHistogram("histogram_Log", da => da
.Field("actionTime")
.Interval("1d")
.Format("dd/MM/YYYY hh:mm:ss")
)
)
)
)
);
Console.WriteLine(Encoding.UTF8.GetString(docs.RequestInformation.Request));
}
public class LogInfoIndexView
{
public string Cluster { get; set; }
public string Server { get; set; }
public string Type { get; set; }
public DateTime ActionTime { get; set; }
}
返回
{
"size": 0,
"aggs": {
"log_query": {
"filter": {
"bool": {
"must": [
{
"term": {
"cluster": "giauht1"
}
},
{
"term": {
"server": "hadoop0"
}
},
{
"term": {
"type": "Warn"
}
},
{
"range": {
"actionTime": {
"lte": "2015-02-24T00:00:00.000",
"gte": "2015-02-01T00:00:00.000"
}
}
}
]
}
},
"aggs": {
"histogram_Log": {
"date_histogram": {
"field": "actionTime",
"interval": "1d",
"format": "dd/MM/YYYY hh:mm:ss"
}
}
}
}
}
}
编辑:
在回答您的评论时,filtered query filter
and a filter aggregation
之间的区别在于前者在查询阶段开始时将过滤应用于所有文档,并且过滤器通常被缓存,从而提高后续查询的性能过滤器,而后者适用于聚合范围,将当前上下文中的文档过滤到单个桶中。如果您的查询只是为了执行聚合并且您可能 运行 使用相同的过滤器进行聚合,我认为 filtered query filter
应该提供更好的性能。