query text search index on compound text search and timestamp mongodb - mongodb

Question

I have a collection, with a total of documents about 65 million records like this
{
"_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, "
.....
}
I made an compound index on (email, created_int): {"email": text, created_int: -1} for search and filter which name in a range of created_int
But I see that it's poor performance on search.
I tried to use explain on query:
db.getCollection('profile_20201').explain().find({"$text":{"$search":"phamthanhlam17_gmail_com"},
"created_int":{"$lte":1585627013, "$gte":1583035013}}).count()
The explain result is:
{
"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" : {
"$gte" : 1583035013.0
}
}
]
},
"inputStage" : {
"stage" : "IXSCAN",
"keyPattern" : {
"_fts" : "text",
"_ftsx" : 1,
"created_int" : -1.0
},
"indexName" : "email_text_created_int_-1",
"isMultiKey" : true,
"isUnique" : false,
"isSparse" : false,
"isPartial" : false,
"indexVersion" : 2,
"direction" : "backward",
"indexBounds" : {}
}
}
}
}
}
},
"rejectedPlans" : []
},
"serverInfo" : {
},
"ok" : 1.0
}
It's explain stats:
"queryPlanner" : {
"plannerVersion" : 1,
"namespace" : "namespace",
"indexFilterSet" : false,
"parsedQuery" : {
"$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.

Answer 1

Well, it seems like you have created the compound index with text. But in official MongoDB Documentation, it says that:
A compound index can include a text index key in combination with ascending/descending index keys. However, these compound indexes have the following restrictions:
A compound text index cannot include any other special index types, such as multi-key or geospatial index fields.
If the compound text index includes keys preceding the text index key, to perform a $text search, the query predicate must include equality match conditions on the preceding keys. (You are using the range query here)
When creating a compound text index, all text index keys must be listed adjacently in the index specification document.
So, that's the first issue.
Next, I want you to have a look at prefixes, it will help you understand how the compound indexes will be used in your query.
Hope this will help you understand the problem :)