# 聚合管道优化

在本页面

* [投影优化](#projection-optimization)
* [管道序列优化](#pipeline-sequence-optimization)
* [管道聚结优化](#pipeline-coalescence-optimization)
* [例子](#example)

聚合管道操作具有优化阶段，该阶段试图重塑管道以改善性能。

要查看优化程序如何转换特定聚合管道，请在[db.collection.aggregate()](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)方法中包含[explain](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)选项。

优化可能会在不同版本之间发生变化。

## 投影优化

聚合管道可以确定它是否仅需要文档中的字段的子集来获得结果。如果是这样，管道将只使用那些必需的字段，减少通过管道的数据量。

## 管道序列优化

### ($project or $unset or $addFields or $set) + $match 序列优化

对于包含投影阶段([$project](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)或[$unset](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)或[$addFields](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)或[$set](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md))后跟[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段的聚合管道，MongoDB 将[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段中不需要在投影阶段计算的值的任何过滤器移动到投影前的新[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段。

如果聚合管道包含多个投影 and/or [$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段，MongoDB 会为每个[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段执行此优化，将每个[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)过滤器移动到过滤器不依赖的所有投影阶段之前。

考虑以下阶段的管道：

```
{ $addFields: {
    maxTime: { $max: "$times" },
    minTime: { $min: "$times" }
} },
{ $project: {
    _id: 1, name: 1, times: 1, maxTime: 1, minTime: 1,
    avgTime: { $avg: ["$maxTime", "$minTime"] }
} },
{ $match: {
    name: "Joe Schmoe",
    maxTime: { $lt: 20 },
    minTime: { $gt: 5 },
    avgTime: { $gt: 7 }
} }
```

优化器将[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段分成四个单独的过滤器，一个用于[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)查询文档中的每个键。然后优化器将每个筛选器移动到尽可能多的投影阶段之前，根据需要创建新的[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段。鉴于此示例，优化程序生成以下优化管道：

```
{ $match: { name: "Joe Schmoe" } },
{ $addFields: {
                maxTime: { $max: "$times" },
        minTime: { $min: "$times" }
} },
{ $match: { maxTime: { $lt: 20 }, minTime: { $gt: 5 } } },
{ $project: {
        _id: 1, name: 1, times: 1, maxTime: 1, minTime: 1,
        avgTime: { $avg: ["$maxTime", "$minTime"] }
} },
{ $match: { avgTime: { $gt: 7 } } }
```

[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)过滤器`{ avgTime: { $gt: 7 } }`取决于[$project](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段来计算`avgTime`字段。 [$project](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段是此管道中的最后一个投影阶段，因此`avgTime`上的[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)过滤器无法移动。

`maxTime`和`minTime`字段在[$addFields](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段计算，但不依赖于[$project](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段。优化器为这些字段上的过滤器创建了一个新的[$match](https://github.com/mongodb-china/MongoDB-CN-Manual/tree/8490376c81d56eff95abbaddc6ee414b1e1c9705/docs/Aggregation/Aggregation-Pipeline/reference-operator-aggregation-match.html#pipe._S_match)阶段，并将其放在[$project](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段之前。

[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)过滤器`{ name: "Joe Schmoe" }`不使用在[$project](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)或[$addFields](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段计算的任何值，因此它在两个投影阶段之前被移动到新的[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段。

> **\[success] 注意**
>
> 优化后，过滤器`{ name: "Joe Schmoe" }`位于管道开头的[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段。这具有额外的好处，即允许聚合在最初查询集合时在`name`字段上使用索引。有关更多信息，请参见[管道操作符和索引](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)。

### $sort + $match 序列优化

如果序列中带有[$sort](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)后跟[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)，则[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)会移动到[$sort](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)之前，以最大程度的减少要排序的对象的数量。例如，如果管道包含以下阶段：

```
{ $sort: { age : -1 } }, 
{ $match: { status: 'A' } }
```

在优化阶段，优化程序将序列转换为以下内容：

```
{ $match: { status: 'A' } }, 
{ $sort: { age : -1 } }
```

### $redact + $match 序列优化

如果可能，当管道的[$redact](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段紧在[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段之后时，聚合有时可以在[$redact](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段之前添加[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段的一部分。如果添加的[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段位于管道的开头，则聚合可以使用索引以及查询集合来限制进入管道的文档数。有关更多信息，请参见[管道操作符和索引](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)。 例如，如果管道包含以下阶段：

```
{ $redact: { $cond: { if: { $eq: [ "$level", 5 ] }, then: "$$PRUNE", else: "$$DESCEND" } } },
{ $match: { year: 2014, category: { $ne: "Z" } } }
```

优化器可以在[$redact](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段之前添加相同的[$match](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段：

```
{ $match: { year: 2014 } },
{ $redact: { $cond: { if: { $eq: [ "$level", 5 ] }, then: "$$PRUNE", else: "$$DESCEND" } } },
{ $match: { year: 2014, category: { $ne: "Z" } } }
```

### `$project`/ `$unset` + `$skip`序列优化

*3.2版本中的新功能。*

当有一个[`$project`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)或[`$unset`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)之后跟有[`$skip`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)序列时，[`$skip`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md) 会移至[`$project`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)之前。例如，如果管道包括以下阶段：

```
{ $sort: { age : -1 } },
{ $project: { status: 1, name: 1 } },
{ $skip: 5 }
```

在优化阶段，优化器将序列转换为以下内容：

```
{ $sort: { age : -1 } },
{ $skip: 5 },
{ $project: { status: 1, name: 1 } }
```

## 管道聚合优化

如果可能，优化阶段将一个管道阶段合并到其前身。通常，合并发生在任何序列重新排序优化之后。

### `$sort` + `$limit`合并

*Mongodb 4.0版本的改变。*

当一个[`$sort`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)先于[`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)，优化器可以聚结[`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)到[`$sort`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)，如果没有中间阶段的修改文件（例如，使用数[`$unwind`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)，[`$group`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)）。如果有管道阶段会更改和阶段之间的文档数，则MongoDB将不会合并[`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)到 。[`$sort`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)[`$sort`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)[`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)

例如，如果管道包括以下阶段：

```
{ $sort : { age : -1 } },
{ $project : { age : 1, status : 1, name : 1 } },
{ $limit: 5 }
```

在优化阶段，优化器将序列合并为以下内容：

```
{
    "$sort" : {
       "sortKey" : {
          "age" : -1
       },
       "limit" : NumberLong(5)
    }
},
{ "$project" : {
         "age" : 1,
         "status" : 1,
         "name" : 1
  }
}
```

这样，排序操作就可以仅在执行过程中保持最高`n`结果，这`n`是指定的限制，MongoDB仅需要将`n`项目存储在内存中 [\[1\]](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)。有关更多信息，请参见[$ sort运算符和内存](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)。

> 用`$skip`进行序列优化
>
> 如果[`$skip`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)在[`$sort`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md) 和[`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段之间有一个阶段，MongoDB将合并 [`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)到该[`$sort`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段并增加该 [`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)值[`$skip`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)。有关示例，请参见 [$ sort + $ skip + $ limit序列](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)。
>
> [\[1\]](https://docs.mongodb.com/manual/core/aggregation-pipeline-optimization/#id1)当优化仍将适用 `allowDiskUse`是`true`与`n`项目超过 [聚集内存限制](https://docs.mongodb.com/manual/core/aggregation-pipeline-limits/#agg-memory-restrictions)。

### `$limit`+ `$limit`合并

当[`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)紧接着另一个时 [`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)，两个阶段可以合并为一个阶段 [`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)，其中限制量为两个初始限制量中的较小者。例如，管道包含以下序列：

```
{ $limit: 100 },
{ $limit: 10 }
```

然后，第二[`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)级可以聚结到第一 [`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段，并导致在单个[`$limit`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md) 阶段，即限制量`10`是两个初始极限的最小`100`和`10`。

```
{ $limit: 10 }
```

### `$skip`+ `$skip`合并

当[`$skip`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)紧跟另一个[`$skip`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)，这两个阶段可合并成一个单一的[`$skip`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)，其中跳过量为总和的两个初始跳过量。例如，管道包含以下序列：

```
{ $skip: 5 },
{ $skip: 2 }
```

然后，第二[`$skip`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段可以合并到第一 [`$skip`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段，并导致单个[`$skip`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md) 阶段，其中跳过量`7`是两个初始限制`5`和的总和`2`。

```
{ $skip: 7 }
```

### `$match`+ `$match`合并

当一个[`$match`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)紧随另一个紧随其后时 [`$match`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)，这两个阶段可以合并为一个单独 [`$match`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)的条件 [`$and`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)。例如，管道包含以下序列：

```
{ $match: { year: 2014 } },
{ $match: { status: "A" } }
```

然后，第二[`$match`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段可以合并到第一 [`$match`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段，从而形成一个[`$match`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md) 阶段

```
{ $match: { $and: [ { "year" : 2014 }, { "status" : "A" } ] } }
```

### `$lookup` + `$unwind` 合并

*3.2版中的新功能。*

当a [`$unwind`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)立即紧随其后 [`$lookup`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)，并且在 领域[`$unwind`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)运行时，优化程序可以将其合并 到阶段中。这样可以避免创建较大的中间文档。`as`[`$lookup`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)[`$unwind`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)[`$lookup`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)

例如，管道包含以下序列：

```
{
  $lookup: {
    from: "otherCollection",
    as: "resultingArray",
    localField: "x",
    foreignField: "y"
  }
},
{ $unwind: "$resultingArray"}
```

优化器可以将[`$unwind`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段合并为 [`$lookup`](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段。如果使用`explain` 选项运行聚合，则`explain`输出将显示合并阶段：

```
{
  $lookup: {
    from: "otherCollection",
    as: "resultingArray",
    localField: "x",
    foreignField: "y",
    unwinding: { preserveNullAndEmptyArrays: false }
  }
}
```

## 例子

### $limit $skip $limit $skip 序列

止于Mongodb4.0

管道包含一系列交替的[$limit](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)和[$skip](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段：

```
{ $limit: 100 },
{ $skip: 5 },
{ $limit: 10 },
{ $skip: 2 }
```

[$skip $limit 序列优化](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)反转`{ $skip: 5 }`和`{ $limit: 10 }`阶段的位置并增加限制量：

```
{ $limit: 100 },
{ $limit: 15},
{ $skip: 5 },
{ $skip: 2 }
```

然后，优化器将两个[$limit](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段合并为一个[$limit](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段，将两个[$skip](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段合并为一个[$skip](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)阶段。结果序列如下：

```
{ $limit: 15 },
{ $skip: 7 }
```

有关详细信息，请参阅[$limit $limit 合并](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)和[$skip $skip 合并](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)。

> **\[success] 可以看看**
>
> [db.collection.aggregate()](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)中的[说明](/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md)选项

译者：李冠飞

校对：李冠飞


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.mongoing.com/aggregation/aggregation-pipeline/aggregation-pipeline-optimization.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.