Docs Home → Develop Applications → MongoDB Manual

Create a Wildcard Index on All Fields

On this page

About this Task

Before You Begin
Procedure
Results
Learn More

You can create a wildcard index that supports queries on all possible document fields. Wildcard indexes support queries on arbitrary or unknown field names.

To create a wildcard index on all fields (excluding _id), use the wildcard specifier ($**) as the index key:

db.<collection>.createIndex( { "$**": <sortOrder> } )

About this Task

Only use wildcard indexes when the fields you want to index are unknown or may change. Wildcard indexes don't perform as well as targeted indexes on specific fields. If your collection contains arbitrary field names that prevent targeted indexes, consider remodeling your schema to have consistent field names. To learn more about targeted indexes, see Create Indexes to Support Your Queries.

Before You Begin

Create an artwork collection that contains the following documents:

db.artwork.insertMany( [
   {
      "name": "The Scream",
      "artist": "Edvard Munch",
      "style": "modern",
      "themes": [ "humanity", "horror" ]
   },
   {
      "name": "Acrobats",
      "artist": {
         "name": "Raoul Dufy",
         "nationality": "French",
         "yearBorn": 1877
      },
      "originalTitle": "Les acrobates",
      "dimensions": [ 65, 49 ]
   },
   {
      "name": "The Thinker",
      "type": "sculpture",
      "materials": [ "bronze" ],
      "year": 1904
   }
] )

Each document contains details about the artwork. The field names vary between documents depending on the information available about the piece.

Procedure

The following operation creates a wildcard index on all document fields in the artwork collection (excluding _id):

db.artwork.createIndex( { "$**" : 1 } )

Results

This index supports single-field queries on any field in the collection. If a document contains an embedded document or array, the wildcard index traverses the document or array and stores the value for all fields in the document or array.

For example, the index supports the following queries:

Query:
```
db.artwork.find( { "style": "modern" } )
```
Output:
```
[
   {
      _id: ObjectId("6352c401b1fac2ee2e957f09"),
      name: 'The Scream',
      artist: 'Edvard Munch',
      style: 'modern',
      themes: [ 'humanity', 'horror' ]
   }
]
```

Query:

db.artwork.find( { "artist.nationality": "French" } )

Output:

[
   {
      _id: ObjectId("6352c525b1fac2ee2e957f0d"),
      name: 'Acrobats',
      artist: { name: 'Raoul Dufy', nationality: 'French', yearBorn: 1877 },
      originalTitle: 'Les acrobates',
      dimensions: [ 65, 49 ]
   }
]

Query:

db.artwork.find( { "materials": "bronze" } )

Output:

[
   {
      _id: ObjectId("6352c387b1fac2ee2e957f08"),
      name: 'The Thinker',
      type: 'sculpture',
      materials: [ 'bronze' ],
      year: 1904
   }
]

Learn More

To learn how to create a wildcard index that projects specific fields to cover, see the following pages:

To learn more about behaviors for wildcard indexes, see:

← Include or Exclude Fields in a Wildcard Index

Compound Wildcard Indexes →