Mongo Aggregates and How to explain Aggregate queries

Updated on: April 4, 2020 · 15 mins read
Categories: database   | nosql   | mongo   | node |

Mongo aggregates with explain
Today we are going to talk about Mongo Aggregates. A framework that helps us to run complex queries in Mongo DB. One of the best things that happened to Mongo.

Let’s start by pulling out a few differences between the normal( SQL) and Mongo(NoSQL) database.

Difference between SQL and NoSQL database

Mongo belongs to one of those NoSQL databases which disrupted the internet a few years ago. Everyone in the industry was talking about them. Everyone wanted to move their stack to these flexible databases.

Everyone was talking about how the data needs to move to that direction and so on.

As the hype began to settle, people started realizing the movement of the stack will help only if they implement it correctly and for most of them, the shift wasn’t even necessary.

NoSQL is a wide term and that is used by a variety of databases. Mongo is also one of them, others being Cassandra, Apache Spark and many more.

MongoDB is a document-based, distributed database. In production, people tend to run it with 3 replicas. 1 is the master and the other two being the slaves. This provides redundancy and high availability.

You can configure these to follow any guidelines but by default reads and writes are handled by the primary replica and the new data is moved on to the replica sets on each database write.

As the mongo doesn’t have a well-defined schema, it’s pretty hard to make queries from the data.

Mongoose is a library that can help you with this. It provides a lot of benefits, like creating hooks and indexes easily on the collections.

BTW, SQL table equivalent in Mongo is known as a collection and the row equivalent is known as a document.

Read more about the differences at GeeksforGeeks.

Mongo Aggregates

Mongo DB
Mongo DB aggregates make it easier to query data from different collections. It involves things like matching, getting data from other collections, selecting fields and many more.

An aggregate is basically a set of operations where the result of one operation is passed on as an input to the next operation where each operation is applying some logic to get further insight from the passed data.

Let’s discuss a few of these aggregate queries.

Starting out a Mongo Aggregate

You can start the aggregate using the following code.

db.collection.aggregate([aggregate pipeline commands], options)
where the collection is the name of the collection on which aggregate is applied and db is the instance of the connected DB object.

Following are the commands that you can use in the aggregate pipeline.

Match in Mongo Aggregate

The match query can be called as the where query in SQL terms. You tell the aggregate to get the data that follows the given condition. It is recommended to keep the match query as soon as possible in the pipeline.

This will reduce the number of documents being returned from the query. It is also a good option to index the fields on which you run the match query to return the results faster.

For example: In a student database you can make this query as follows

{ $match: { "roll_number": 901 }}
or

{ $match: { "class": 5 }}
Corresponding SQL query for this would be

select * from student where class = 5;
This query will return all the documents which satisfy the given query.

For all further parts of the pipeline, you can keep adding it to the main array of the pipeline.

Note: You can make use of the fields by appending a $ to the name of the fields, whenever you want to use them.

Use match as early as possible

You should use $match as early as possible to make use of indexes.

$match operations use suitable indexes to scan only the matching documents in a collection. When possible, place $match operators at the beginning of the pipeline.

Always try to explain the $match part of your queries and prefer to create compound indexes according to your queries.

Match vs find in mongo

$match produces the same results as of find query that is why it’s hard to understand the difference between the aggregates and normal queries if you start by studying match queries but please tolerate me for some more time. You will know what we are talking about in some time.

Match vs find in mongo

Project in Mongo Aggregate

Project is the part where you tell the query which keys to pick from the given document.

{
    $project: {
        student_name: 1,
        student_age: '$age'
    }
}
This will pick up only the fields with value 1 in the document. It is important to reduce the data being transferred to the next part of the pipeline.

It can also be used to change the name of the field. This is the SELECT equivalent of SQL commands.

select student_name, student_age from students;
_id field is added by default in the result. You can use _id: 0, if you don’t want to include that field.

Project in mongo
I hope you have started to understand the usage of aggregates in a better way.

Grouping in Mongo Aggregate

Group command is used to group different things together, for example, if you want to calculate the sum of the age of students in different standards, the query will look something like this.

db.students.aggregate([
    { $group: { _id: "$class", total: { $sum: "$age" } } }
])
Here is the link to the mongo query. You can use all different types of aggregate queries like average, min and max.

Lookup other collections in Mongo Aggregate

Lookup is one of the most important aggregate queries which you can use. This allows us to join different collections together and get data from the other collections as well.

The simplest implementation of $lookup is as follows.

{
    $lookup: {
        from: <collection to join>,
        localField: <field from the input documents>,
        foreignField: <field from the documents of the "from" collection>,
        as: <output array field>
    }
}
Here is the query of the lookup.

The lookup can be used with the pipeline as well. With this type of lookup, you can apply a check and tell the pipeline, when the given lookup will run.

{
      $lookup:
         {
           from: "students",
           let: { studentId: "$_id", age: "$age" },
           pipeline: [
              { $match:
                 { $expr:
                     { $eq: [ "$student_id",  "$$studentId" ] }
                 }
              },
              { $project: { student_id: 1, age: 1 } }
           ],
           as: "data"
         }
    }
This is the implementation that you want to use, when you want to run $match on the data being picked from the other collection.

Unwind in Mongo Aggregate

Unwinding is a type of operation in which deconstructs an array field from the input documents to output a document for each element.

For example, consider this document:

[
  {
    "collection": "collection",
    "count": 10,
    "content": [
      {
        "k": {
          "type": "int",
          "minInt": 0,
          "maxInt": 10
        },
        "a": {
          "type": "int",
          "minInt": 0,
          "maxInt": 10
        },
        "b": {
          "type": "int",
          "minInt": 0,
          "maxInt": 10
        },
        
      },
      {}
    ]
  }
]
On applying $unwind,

db.collection.aggregate([
  {
    "$unwind": {
      "path": "$content",
      "preserveNullAndEmptyArrays": true
    }
  },
  {
    "$project": {
      "_id": 0,
      "content": 1
    }
  }
])
We will get the following result,

[
  {
    "content": {
      "a": {
        "maxInt": 10,
        "minInt": 0,
        "type": "int"
      },
      "b": {
        "maxInt": 10,
        "minInt": 0,
        "type": "int"
      },
      "k": {
        "maxInt": 10,
        "minInt": 0,
        "type": "int"
      }
    }
  },
  {
    "content": {}
  }
]
Check this on Mongo playground.

Mongo Aggregation Pipeline

When data is passed on from multiple such aggregation layers of match, lookup, project etc, it is called Mongo Aggregation Pipeline. (Just a fancy term)

How to Explain mongo aggregate queries using Mongodb Explain command

To explain the queries, you will have to use the options in the aggregate to find the way in which queries are run.

db.getCollection("author").explain().aggregate([
	{
		$match: {
		    "email" : "[email protected]"
		}
	}
])
or from shell,

db.author.explain().aggregate([
	{
		$match: {
		    "email" : "[email protected]"
		}
	}
])
This will generate a simple output like this.

{
    "stages" : [
        {
            "$cursor" : {
                "query" : {
                    "email" : "[email protected]"
                },
                "queryPlanner" : {
                    "plannerVersion" : 1.0,
                    "namespace" : "db_name.author",
                    "indexFilterSet" : false,
                    "parsedQuery" : {
                        "email" : {
                            "$eq" : "[email protected]"
                        }
                    },
                    "winningPlan" : {
                        "stage" : "COLLSCAN",
                        "filter" : {
                            "email" : {
                                "$eq" : "[email protected]"
                            }
                        },
                        "direction" : "forward"
                    },
                    "rejectedPlans" : [

                    ]
                }
            }
        }
    ],
    "ok" : 1.0
}
winningplan contains an object which tells us more about the winning plan which was used to run the query and queryPlanner contains an Array of plans which were tried. Mongo chooses the best plans and uses it for running the queries.

If you use the explain with executionStats it will give you things like docs Returned and docs Examined which can be helpful in finding the best suitable index for your collection.

db.getCollection("author").explain("executionStats").aggregate([
	{
		$match: {
		    "email" : "[email protected]"
		}
	}
])
Explain mongo aggregate queries
In the following example, the query is examining all 5 rows before giving the result.

Mongodb explain
We can reduce that by indexing the item field.

Create index in mongodb
Now when we apply the executionStats command we get a better docs examined ratio.

execution stats after index

Conclusion

There are a few more options to choose from. Aggregate is one of the most important parts of the Mongo Database.

If you are dealing with this Database daily, then it would be useful to know a little about it as well.

Thanks for stopping by, Let me know if you want to know about something else as well. I love to write about tech topics.

Also, let me know if I have made any mistake in this post.

Note: The internal performance optimizer of the mongo aggregate optimizes the queries accordingly to make them as fast as possible.

Please share your Feedback:

Did you enjoy reading or think it can be improved? Don’t forget to leave your thoughts in the comments section below! If you liked this article, please share it with your friends, and read a few more!

We don't share your details with others