我有一个馆藏,总共有大约6500万条这样的记录
{
"_id" : ObjectId("5e0b814660da38d499ecf178"),
"brands" : null,
"client_id" : null,
"code_co_owner" : ",7359562, ",
"code_segment" : "7359562",
"core" : "",
"created" : "01-01-2020",
"created_full" : "01-01-2020 00:00:27",
"created_int" : NumberLong(1577811627),
"email" : ",phamthanhlam17_gmail_com, "
.....
}
我在(email,created_int)上做了一个复合索引:{“ email”:文本,created_int:-1}用于搜索和过滤在created_int范围内的名称 但我发现搜索效果不佳。
我试图在查询中使用解释:
db.getCollection('profile_20201').explain().find({"$text":{"$search":"phamthanhlam17_gmail_com"},
"created_int":{"$lte":1585627013, "$gte":1583035013}}).count()
解释结果是:
{
"queryPlanner" : {
"plannerVersion" : 1,
"indexFilterSet" : false,
"parsedQuery" : {
"$and" : [
{
"created_int" : {
"$lte" : 1585627013.0
}
},
{
"created_int" : {
"$gte" : 1583035013.0
}
},
{
"$text" : {
"$search" : "phamthanhlam17_gmail_com",
"$language" : "english",
"$caseSensitive" : false,
"$diacriticSensitive" : false
}
}
]
},
"winningPlan" : {
"stage" : "COUNT",
"inputStage" : {
"stage" : "TEXT",
"indexPrefix" : {},
"indexName" : "email_text_created_int_-1",
"parsedTextQuery" : {
"terms" : [
"phamthanhlam17_gmail_com"
],
"negatedTerms" : [],
"phrases" : [],
"negatedPhrases" : []
},
"textIndexVersion" : 3,
"inputStage" : {
"stage" : "TEXT_MATCH",
"inputStage" : {
"stage" : "FETCH",
"inputStage" : {
"stage" : "OR",
"filter" : {
"$and" : [
{
"created_int" : {
"$lte" : 1585627013.0
}
},
{
"created_int" : {
"$gte" : 1583035013.0
}
}
]
},
"inputStage" : {
"stage" : "IXSCAN",
"keyPattern" : {
"_fts" : "text",
"_ftsx" : 1,
"created_int" : -1.0
},
"indexName" : "email_text_created_int_-1",
"isMultiKey" : true,
"isUnique" : false,
"isSparse" : false,
"isPartial" : false,
"indexVersion" : 2,
"direction" : "backward",
"indexBounds" : {}
}
}
}
}
}
},
"rejectedPlans" : []
},
"serverInfo" : {
},
"ok" : 1.0
}
解释统计信息:
"queryPlanner" : {
"plannerVersion" : 1,
"namespace" : "namespace",
"indexFilterSet" : false,
"parsedQuery" : {
"$and" : [
{
"created_int" : {
"$lte" : 1585627013.0
}
},
{
"created_int" : {
"$gte" : 1583035013.0
}
},
{
"$text" : {
"$search" : "phamthanhlam17_gmail_com",
"$language" : "english",
"$caseSensitive" : false,
"$diacriticSensitive" : false
}
}
]
},
"winningPlan" : {
"stage" : "COUNT",
"inputStage" : {
"stage" : "TEXT",
"indexPrefix" : {},
"indexName" : "email_text_created_int_-1",
"parsedTextQuery" : {
"terms" : [
"phamthanhlam17_gmail_com"
],
"negatedTerms" : [],
"phrases" : [],
"negatedPhrases" : []
},
"textIndexVersion" : 3,
"inputStage" : {
"stage" : "TEXT_MATCH",
"inputStage" : {
"stage" : "FETCH",
"inputStage" : {
"stage" : "OR",
"filter" : {
"$and" : [
{
"created_int" : {
"$lte" : 1585627013.0
}
},
{
"created_int" : {
"$gte" : 1583035013.0
}
}
]
},
"inputStage" : {
"stage" : "IXSCAN",
"keyPattern" : {
"_fts" : "text",
"_ftsx" : 1,
"created_int" : -1.0
},
"indexName" : "email_text_created_int_-1",
"isMultiKey" : true,
"isUnique" : false,
"isSparse" : false,
"isPartial" : false,
"indexVersion" : 2,
"direction" : "backward",
"indexBounds" : {}
}
}
}
}
}
},
"rejectedPlans" : []
},
"executionStats" : {
"executionSuccess" : true,
"nReturned" : 0,
"executionTimeMillis" : 1499057,
"totalKeysExamined" : 72544123,
"totalDocsExamined" : 39448083,
"executionStages" : {
"stage" : "COUNT",
"nReturned" : 0,
"executionTimeMillisEstimate" : 1483861,
"works" : 72544124,
"advanced" : 0,
"needTime" : 72544123,
"needYield" : 0,
"saveState" : 578233,
"restoreState" : 578233,
"isEOF" : 1,
"invalidates" : 0,
"nCounted" : 39448083,
"nSkipped" : 0,
"inputStage" : {
"stage" : "TEXT",
"nReturned" : 39448083,
"executionTimeMillisEstimate" : 1475831,
"works" : 72544124,
"advanced" : 39448083,
"needTime" : 33096040,
"needYield" : 0,
"saveState" : 578233,
"restoreState" : 578233,
"isEOF" : 1,
"invalidates" : 0,
"indexPrefix" : {},
"indexName" : "email_text_created_int_-1",
"parsedTextQuery" : {
"terms" : [
"phamthanhlam17_gmail_com"
],
"negatedTerms" : [],
"phrases" : [],
"negatedPhrases" : []
},
"textIndexVersion" : 3,
"inputStage" : {
"stage" : "TEXT_MATCH",
"nReturned" : 39448083,
"executionTimeMillisEstimate" : 1473041,
"works" : 72544124,
"advanced" : 39448083,
"needTime" : 33096040,
"needYield" : 0,
"saveState" : 578233,
"restoreState" : 578233,
"isEOF" : 1,
"invalidates" : 0,
"docsRejected" : 0,
"inputStage" : {
"stage" : "FETCH",
"nReturned" : 39448083,
"executionTimeMillisEstimate" : 1465951,
"works" : 72544124,
"advanced" : 39448083,
"needTime" : 33096040,
"needYield" : 0,
"saveState" : 578233,
"restoreState" : 578233,
"isEOF" : 1,
"invalidates" : 0,
"docsExamined" : 39448083,
"alreadyHasObj" : 0,
"inputStage" : {
"stage" : "OR",
"filter" : {
"$and" : [
{
"created_int" : {
"$lte" : 1585627013.0
}
},
{
"created_int" : {
"$gte" : 1583035013.0
}
}
]
},
"nReturned" : 39448083,
"executionTimeMillisEstimate" : 439664,
"works" : 72544124,
"advanced" : 39448083,
"needTime" : 33096040,
"needYield" : 0,
"saveState" : 578233,
"restoreState" : 578233,
"isEOF" : 1,
"invalidates" : 0,
"dupsTested" : 72544123,
"dupsDropped" : 0,
"recordIdsForgotten" : 0,
"inputStage" : {
"stage" : "IXSCAN",
"nReturned" : 72544123,
"executionTimeMillisEstimate" : 291188,
"works" : 72544124,
"advanced" : 72544123,
"needTime" : 0,
"needYield" : 0,
"saveState" : 578233,
"restoreState" : 578233,
"isEOF" : 1,
"invalidates" : 0,
"keyPattern" : {
"_fts" : "text",
"_ftsx" : 1,
"created_int" : -1.0
},
"indexName" : "email_text_created_int_-1",
"isMultiKey" : true,
"isUnique" : false,
"isSparse" : false,
"isPartial" : false,
"indexVersion" : 2,
"direction" : "backward",
"indexBounds" : {},
"keysExamined" : 72544123,
"seeks" : 1,
"dupsTested" : 72544123,
"dupsDropped" : 0,
"seenInvalidated" : 0
}
}
}
}
}
}
},
"serverInfo" : {
},
"ok" : 1.0
}```
So, is the index is cover the query?
Or which index will give me better performance for this problem?
Thank you.
答案 0 :(得分:1)
好吧,看来您已经用text
创建了复合索引。但是在官方的MongoDB Documentation中,它表示:
复合索引可以包含文本索引键和升/降索引键。但是,这些复合索引具有以下限制:
所以,这是第一个问题。
接下来,我希望您看看prefixes,它将帮助您了解如何在查询中使用复合索引。
希望这可以帮助您理解问题:)