Docs Menu

Docs HomeDevelop ApplicationsMongoDB Manual

Explain Results

On this page

  • Explain Output
  • queryPlanner
  • executionStats
  • serverInfo
  • 3.0 Format Change
  • Collection Scan vs. Index Use
  • Covered Queries
  • Index Intersection
  • $or Expression
  • Sort Stage

To return information on query plans and execution statistics of the query plans, MongoDB provides:

The explain results present the query plans as a tree of stages.

"winningPlan" : {
"stage" : <STAGE1>,
...
"inputStage" : {
"stage" : <STAGE2>,
...
"inputStage" : {
"stage" : <STAGE3>,
...
}
}
},

Each stage passes its results (i.e. documents or index keys) to the parent node. The leaf nodes access the collection or the indices. The internal nodes manipulate the documents or the index keys that result from the child nodes. The root node is the final stage from which MongoDB derives the result set.

Stages are descriptive of the operation; e.g.

  • COLLSCAN for a collection scan

  • IXSCAN for scanning index keys

  • FETCH for retrieving documents

  • SHARD_MERGE for merging results from shards

  • SHARDING_FILTER for filtering out orphan documents from shards

The following sections presents a list of some key fields returned by the explain operation.

Note

  • The list of fields is not meant to be exhaustive, but is meant to highlight some key field changes from earlier versions of explain.

  • The output format is subject to change between releases.

queryPlanner information details the plan selected by the query optimizer.

explain.queryPlanner

Contains information on the selection of the query plan by the query optimizer.

explain.queryPlanner.namespace

A string that specifies the namespace with the names of the database and the collection accessed by the query. The namespace has the format <database>.<collection>.

explain.queryPlanner.indexFilterSet

A boolean that specifies whether MongoDB applied an index filter for the query shape.

explain.queryPlanner.queryHash

A hexadecimal string that represents the hash of the query shape and is dependent only on the query shapes. queryHash can help identify slow queries (including the query filter of write operations) with the same query shape.

Note

As with any hash function, two different query shapes may result in the same hash value. However, the occurrence of hash collisions between different query shapes is unlikely.

For more information on queryHash and planCacheKey, see queryHash and planCacheKey.

New in version 4.2.

explain.queryPlanner.planCacheKey

A hash of the key for the plan cache entry associated with the query.

Unlike the queryHash, the planCacheKey is a function of both the query shape and the currently available indexes for that shape. That is, if indexes that can support the query shape are added/dropped, the planCacheKey value may change whereas the queryHash value would not change.

For more information on queryHash and planCacheKey, see queryHash and planCacheKey.

New in version 4.2.

explain.queryPlanner.optimizedPipeline

A boolean that indicates that the entire aggregation pipeline operation was optimized away, and instead, fulfilled by a tree of query plan execution stages.

For example, starting in MongodB 4.2, the following aggregation operation can be fulfilled by the tree of query plan execution rather than using the aggregation pipeline.

db.example.aggregate([ { $match: { someFlag: true } } ] )

The field is only present if the value is true and only applies to explain on aggregation pipeline operations. When true, because the pipeline was optimized away, no aggregation stage information appears in the output.

New in version 4.2.

explain.queryPlanner.winningPlan

A document that details the plan selected by the query optimizer. MongoDB presents the plan as a tree of stages; i.e. a stage can have an inputStage or, if the stage has multiple child stages, inputStages.

explain.queryPlanner.winningPlan.stage

A string that denotes the name of the stage.

Each stage consists of information specific to the stage. For instance, an IXSCAN stage will include the index bounds along with other data specific to the index scan. If a stage has a child stage or multiple child stages, the stage will have an inputStage or inputStages.

explain.queryPlanner.winningPlan.inputStage

A document that describes the child stage, which provides the documents or index keys to its parent. The field is present if the parent stage has only one child.

explain.queryPlanner.winningPlan.inputStages

An array of documents describing the child stages. Child stages provide the documents or index keys to the parent stage. The field is present if the parent stage has multiple child nodes. For example, stages for $or expressions or index intersection consume input from multiple sources.

explain.queryPlanner.rejectedPlans

Array of candidate plans considered and rejected by the query optimizer. The array can be empty if there were no other candidate plans.

The returned executionStats information details the execution of the winning plan. In order to include executionStats in the results, you must run the explain in either:

explain.executionStats

Contains statistics that describe the completed query execution for the winning plan. For write operations, completed query execution refers to the modifications that would be performed, but does not apply the modifications to the database.

explain.executionStats.nReturned

Number of documents returned by the winning query plan. nReturned corresponds to the n field returned by cursor.explain() in earlier versions of MongoDB.

explain.executionStats.executionTimeMillis

Total time in milliseconds required for query plan selection and query execution. It includes the time it takes to run the trial phase part of the plan selection process, but does not include the network time to transmit the data back to the client.

The time reported by explain.executionStats.executionTimeMillis is not necessarily representative of actual query time. During steady state operations (when the query plan is cached), or when using cursor.hint() with cursor.explain(), MongoDB bypasses the plan selection process, resulting in a faster actual time, leading to a lower explain.executionStats.executionTimeMillis value.

explain.executionStats.totalKeysExamined

Number of index entries scanned. totalKeysExamined corresponds to the nscanned field returned by cursor.explain() in earlier versions of MongoDB.

explain.executionStats.totalDocsExamined

Number of documents examined during query execution. Common query execution stages that examine documents are COLLSCAN and FETCH.

Note

totalDocsExamined refers to the total number of documents examined and not to the number of documents returned. For example, a stage can examine a document in order to apply a filter. If the document is filtered out, then it has been examined but will not be returned as part of the query result set.

If a document is examined multiple times during query execution, totalDocsExamined counts each examination. That is, totalDocsExamined is not a count of the total number of unique documents examined.

explain.executionStats.executionStages

Details the completed execution of the winning plan as a tree of stages; i.e. a stage can have an inputStage or multiple inputStages.

Each stage consists of execution information specific to the stage.

explain.executionStats.executionStages.executionTimeMillisEstimate

The estimated amount of time in milliseconds for query execution.

explain.executionStats.executionStages.works

Specifies the number of "work units" performed by the query execution stage. Query execution divides its work into small units. A "work unit" might consist of examining a single index key, fetching a single document from the collection, applying a projection to a single document, or doing a piece of internal bookkeeping.

explain.executionStats.executionStages.advanced

The number of intermediate results returned, or advanced, by this stage to its parent stage.

explain.executionStats.executionStages.needTime

The number of work cycles that did not advance an intermediate result to its parent stage (see explain.executionStats.executionStages.advanced). For instance, an index scan stage may spend a work cycle seeking to a new position in the index as opposed to returning an index key; this work cycle would count towards explain.executionStats.executionStages.needTime rather than explain.executionStats.executionStages.advanced.

explain.executionStats.executionStages.needYield

The number of times that the storage layer requested that the query stage suspend processing and yield its locks.

explain.executionStats.executionStages.saveState

The number of times that the query stage suspended processing and saved its current execution state, for example in preparation for yielding its locks.

explain.executionStats.executionStages.restoreState

The number of times that the query stage restored a saved execution state, for example after recovering locks that it had previously yielded.

explain.executionStats.executionStages.isEOF

Specifies whether the execution stage has reached end of stream:

  • If true or 1, the execution stage has reached end-of-stream.

  • If false or 0, the stage may still have results to return. For example, consider a query with a limit whose execution stages consists of a LIMIT stage with an input stage of IXSCAN for the query. If the query returns more than the specified limit, the LIMIT stage will report isEOF: 1, but its underlying IXSCAN stage will report isEOF: 0.

explain.executionStats.executionStages.inputStage.keysExamined

For query execution stages that scan an index (e.g. IXSCAN), keysExamined is the total number of in-bounds and out-of-bounds keys that are examined in the process of the index scan. If the index scan consists of a single contiguous range of keys, only in-bounds keys need to be examined. If the index bounds consists of several key ranges, the index scan execution process may examine out-of-bounds keys in order to skip from the end of one range to the beginning of the next.

Consider the following example, where there is an index of field x and the collection contains 100 documents with x values 1 through 100:

db.keys.find( { x : { $in : [ 3, 4, 50, 74, 75, 90 ] } } ).explain( "executionStats" )

The query will scan keys 3 and 4. It will then scan the key 5, detect that it is out-of-bounds, and skip to the next key 50.

Continuing this process, the query scans keys 3, 4, 5, 50, 51, 74, 75, 76, 90, and 91. Keys 5, 51, 76, and 91 are out-of-bounds keys that are still examined. The value of keysExamined is 10.

explain.executionStats.executionStages.inputStage.docsExamined

Specifies the number of documents scanned during the query execution stage.

Present for the COLLSCAN stage, as well as for stages that retrieve documents from the collection (e.g. FETCH)

explain.executionStats.executionStages.inputStage.seeks

New in version 3.4: For index scan (IXSCAN) stages only.

The number of times that we had to seek the index cursor to a new position in order to complete the index scan.

explain.executionStats.allPlansExecution

Contains partial execution information captured during the plan selection phase for both the winning and rejected plans. The field is present only if explain runs in allPlansExecution verbosity mode.

Starting in MongoDB 3.0, the format and fields of the explain results have changed from previous versions. The following lists some key differences.

If the query planner selects a collection scan, the explain result includes a COLLSCAN stage.

If the query planner selects an index, the explain result includes a IXSCAN stage. The stage includes information such as the index key pattern, direction of traversal, and index bounds.

In previous versions of MongoDB, cursor.explain() returned the cursor field with the value of:

  • BasicCursor for collection scans, and

  • BtreeCursor <index name> [<direction>] for index scans.

For more information on execution statistics of collection scans versus index scans, see Analyze Query Performance.

When an index covers a query, MongoDB can both match the query conditions and return the results using only the index keys; i.e. MongoDB does not need to examine documents from the collection to return the results.

When an index covers a query, the explain result has an IXSCAN stage that is not a descendant of a FETCH stage, and in the executionStats, the totalDocsExamined is 0.

In earlier versions of MongoDB, cursor.explain() returned the indexOnly field to indicate whether the index covered a query.

For an index intersection plan, the result will include either an AND_SORTED stage or an AND_HASH stage with an inputStages array that details the indexes; e.g.:

{
"stage" : "AND_SORTED",
"inputStages" : [
{
"stage" : "IXSCAN",
...
},
{
"stage" : "IXSCAN",
...
}
]
}

In previous versions of MongoDB, cursor.explain() returned the cursor field with the value of Complex Plan for index intersections.

If MongoDB uses indexes for an $or expression, the result will include the OR stage with an inputStages array that details the indexes; e.g.:

{
"stage" : "OR",
"inputStages" : [
{
"stage" : "IXSCAN",
...
},
{
"stage" : "IXSCAN",
...
},
...
]
}

In previous versions of MongoDB, cursor.explain() returned the clauses array that detailed the indexes.

If MongoDB cannot use an index or indexes to obtain the sort order, the results include a SORT stage indicating a blocking sort operation. Blocking sorts do not block concurrent operations on the collection or database. The name refers to the requirement that the SORT stage reads all input documents before returning any output documents, blocking the flow of data for that specific query.

If MongoDB requires using more than 100 megabytes of system memory for the blocking sort operation, MongoDB returns an error unless the query specifies cursor.allowDiskUse() (New in MongoDB 4.4). allowDiskUse() allows MongoDB to use temporary files on disk to store data exceeding the 100 megabyte system memory limit while processing a blocking sort operation. If the explain plan does not contain an explicit SORT stage, then MongoDB can use an index to obtain the sort order.

Prior to MongoDB 3.0, cursor.explain() returned the scanAndOrder field to specify whether MongoDB could use the index order to return sorted results.

←  Write Operation PerformanceAnalyze Query Performance →