Reduce Number of Collections

Overview

Collections are groupings of MongoDB documents, similar to an RDBMS table. A collection exists within a single database.

Even if a collection does not contain any documents, it still comes with a resource cost in the form of an undroppable default _id index. Although this index does not take up much space on its own (especially for small collections) if you have thousands of collections these indexes can add up in resources and strain your database allocation.

If your deployment contains unnecessary or an increasing number of collections, you should consider restructuring your data to reduce the number of collections and ultimately reduce the resource requirements of your application.

Example

Consider a temperatures database that stores collections of temperature readings obtained from a sensor. The sensor takes readings every half hour from 10 AM to 10 PM. Each day's readings are stored in a separate collection, named by the reading date:

// temperatures.march-09-2020
{
  "_id": 1,
  "timestamp": "2020-03-09T010:00:00Z",
  "temperature": 29
}
{
  "_id": 2,
  "timestamp": "2020-03-09T010:30:00Z",
  "temperature": 30
}
...
{
  "_id": 25,
  "timestamp": "2020-03-09T022:00:00Z",
  "temperature": 26
}

// temperatures.march-10-2020
{
  "_id": 1,
  "timestamp": "2020-03-10T010:00:00Z",
  "temperature": 30
}
{
  "_id": 2,
  "timestamp": "2020-03-10T010:30:00Z",
  "temperature": 32
}
...
{
  "_id": 25,
  "timestamp": "2020-03-10T022:00:00Z",
  "temperature": 28
}

With each passing day, the number of collections in the database increments. Since the number of collections is unbounded, there is an ever-growing need from the database to maintain these collections and their corresponding indexes. If the database eventually reaches a point where it is managing thousands of collections and indexes, it may result in performance degradation.

Additionally, this approach does not easily facilitate queries across multiple days. To query data from multiple days to obtain temperature trends over longer periods of time, you would need to perform a $lookup operation, which is not as performant as querying data in the same collection.

// temperatures.readings
{
  "_id": ISODate("2020-03-09"),
  "readings": [
    {
      "timestamp": "2020-03-09T010:00:00Z",
      "temperature": 29
    },
    {
      "timestamp": "2020-03-09T010:30:00Z",
      "temperature": 30
    },
    ...
    {
      "timestamp": "2020-03-09T022:00:00Z",
      "temperature": 26
    }
  ]
}
{
  "_id": ISODate("2020-03-10"),
  "readings": [
    {
      "timestamp": "2020-03-10T010:00:00Z",
      "temperature": 30
    },
    {
      "timestamp": "2020-03-10T010:30:00Z",
      "temperature": 32
    },
    ...
    {
      "timestamp": "2020-03-10T022:00:00Z",
      "temperature": 28
    }
  ]
}

How to Check for Unnecessary Collections

db.getCollectionNames().length

Learn More

• To learn more about Data Modeling in MongoDB and the flexible schema model, see Data Modeling Introduction.

• To learn more about databases and collections, see Databases and Collections.

• To learn more about embedding related data in a single document, see Embedded Data Models