Index types in Cloud Firestore

Indexes are an important factor in the performance of a database. Much like the index of a book which maps topics in a book to page numbers, a database index maps the items in a database to their locations in the database. When you query a database, the database can use an index to quickly identify the locations of the items that you requested.

This page describes the two types of indexes that Cloud Firestore uses, single-field indexes and composite indexes.

Index definition and structure

An index is defined on a list of fields of a given document, with a corresponding index mode for each field.

An index contains an entry for every field named in the index definition. The index includes all documents that are the potential results for queries based on the index. A document is included in the index only if it has an indexed value set for every field used in the index. If the index definition refers to a field for which the document has no value set, that document won't appear in the index. In this case, the document will never be returned as a result for any query based on the index.

The composite index is sorted by field values, in the order specified in the index definition.

An index behind every query

If no index exists for a query, most databases crawl through their contents item by item, a slow process that slows down even more as the database grows. Cloud Firestore guarantees high query performance by using indexes for all queries. As a result, query performance depends on the size of the result set and not on the number of items in the database.

Less index management, more app development

Cloud Firestore includes features that reduce the amount of time that you need to spend on index management. The indexes required for the most basic queries are automatically created for you. As you use and test your app, Cloud Firestore helps you identify and create additional indexes that your app requires.

Index types

Cloud Firestore uses two types of indexes: single-field and composite. Besides the number of fields indexed, single-field and composite indexes differ in how you manage them.

Single-field indexes

A single-field index stores a sorted mapping of all the documents in a collection that contain a specific field. Each entry in a single-field index records a document's value for a specific field and the location of the document in the database. Cloud Firestore uses these indexes to perform many basic queries. You manage single-field indexes by configuring your database's automatic indexing settings and index exemptions.

Automatic indexing

By default, Cloud Firestore automatically maintains single-field indexes for each field in a document and each subfield in a map. Cloud Firestore uses the following default settings for single-field indexes:

  • For each non-array and non-map field, Cloud Firestore defines two collection-scope single-field indexes, one in ascending mode and one in descending mode.

  • For each map field, Cloud Firestore creates the following:

    • One collection-scope ascending index for each non-array, non-map subfield.
    • One collection-scope descending index for each non-array, non-map subfield.
    • One collection-scope array-contains index for each array subfield.
    • Cloud Firestore recursively indexes each map subfield.
  • For each array field in a document, Cloud Firestore creates and maintains a collection-scope array-contains index.

  • Single-field indexes with collection group scope are not maintained by default.

Single-field index exemptions

You can exempt a field from your automatic indexing settings by creating a single-field index exemption. An indexing exemption overrides the database-wide automatic index settings. An exemption can enable a single-field index that your automatic indexing settings would otherwise disable or disable a single-field index that automatic indexing would otherwise enable. For cases where exemptions can be useful, see the indexing best practices.

Use the * field path value to add collection-level index exemptions on all fields in a collection group. For example, for collection group comments, set the field path to * to match all fields in the comments collection group and disable indexing of all the fields under the collection group. You can then add exemptions to index only the fields required for your queries. Reducing the number of indexed fields reduces storage costs and can improve write performance.

If you create a single-field index exemption for a map field, the map's subfields inherit those settings. You can, however, define single-field index exemptions for specific subfields. If you delete an exemption for a subfield, the subfield will inherit its parent's exemption settings, if they exist, or the database-wide settings if no parent exemptions exist.

To create and manage single-field index exemptions, see Manage indexes.

Composite indexes

A composite index stores a sorted mapping of all the documents in a collection, based on an ordered list of fields to index.

Cloud Firestore uses composite indexes to support queries not already supported by single-field indexes.

Cloud Firestore doesn't automatically create composite indexes like it does for single-field indexes because of the large number of possible field combinations. Instead, Cloud Firestore helps you identify and create required composite indexes as you build your app.

Any time you attempt a query that isn't supported by an index, Cloud Firestore returns an error message with a link that you can follow to create the missing index.

You can also define and manage composite indexes manually by using the console or by using the Firebase CLI. For more on creating and managing composite indexes, see Manage indexes.

Index modes and query scopes

You configure single-field and composite indexes differently, but both require that you configure index modes and query scopes for your indexes.

Index modes

When you define an index, you select an index mode for each indexed field. Each field's index mode supports specific query clauses on that field. You can select from the following index modes:

Index mode Description
Ascending Supports <, <=, ==, >=, >, !=, in, and not-in, query clauses on the field and supports sorting results in ascending order based on this field value.
Descending Supports <, <=, ==, >=, >, !=, in, and not-in query clauses on the field and supports sorting results in descending order based on this field value.
Array‑contains Supports array-contains and array-contains-any query clauses on the field.
Vector Supports FindNearest query clauses on the field.

Query scopes

Each index is scoped to either a collection or a collection group. This is known as the index's query scope:

Collection scope
Cloud Firestore creates indexes with collection scope by default. These indexes support queries that return results from a single collection.

Collection group scope
A collection group includes all collections with the same collection ID. To run a collection group query that returns filtered or ordered results from a collection group, you must create a corresponding index with collection group scope.

Default ordering and the __name__ field

In addition to sorting documents by the index modes specified for each field (ascending or descending) , indexes apply a final sorting by the __name__ field of each document. The value of the __name__ field is set to the full document path. This means that documents in the result set with the same field values are sorted by document path.

By default, the __name__ field is sorted in the same direction of the last sorted field in the index definition. For example:

Collection Fields indexed Query scope
cities name, __name__ Collection
cities state, __name__ Collection
cities country, population, __name__ Collection

To sort results by the non-default __name__ direction, you need to create that index.

Index properties

An index that allows the query to be executed most efficiently is defined by the following properties:

  • Fields used in equality filters
  • Fields used in sort orders
  • Fields used in range and inequality filters (that are not already included in sort orders)
  • Fields used in aggregations (that aren't already included in sort orders and range and inequality filters)

Cloud Firestore computes the results for queries as follows:

  1. Identifies the index corresponding to the query's collection, filter properties, filter operators, and sort orders.
  2. Identifies the index position from which the scanning starts. The start position is prefixed with the query's equality filters and ends with the range and inequality filters on the first orderBy field.
  3. Starts scanning the index, returning each document that satisfies all the filters, until the scanning process does one of the following:
    • Encounters a document that doesn't meet the filter conditions and confirms that any subsequent document will never fully meet the filter conditions.
    • Reaches the end of the index.
    • Collects the maximum number of results requested by the query.

Indexing example

By automatically creating single-field indexes for you, Cloud Firestore allows your application to quickly support the most basic database queries. Single-field indexes allow you to perform simple queries based on field values and the comparators <, <=, ==, >=, >, and in. For array fields, they allow you to perform array-contains and array-contains-any queries.

To illustrate, examine the following examples from the point of view of index creation. The following snippet creates a few city documents in a cities collection and sets name, state, country, capital, population, and tags fields for each document:

Web
var citiesRef = db.collection("cities");

citiesRef.doc("SF").set({
    name: "San Francisco", state: "CA", country: "USA",
    capital: false, population: 860000,
    regions: ["west_coast", "norcal"] });
citiesRef.doc("LA").set({
    name: "Los Angeles", state: "CA", country: "USA",
    capital: false, population: 3900000,
    regions: ["west_coast", "socal"] });
citiesRef.doc("DC").set({
    name: "Washington, D.C.", state: null, country: "USA",
    capital: true, population: 680000,
    regions: ["east_coast"] });
citiesRef.doc("TOK").set({
    name: "Tokyo", state: null, country: "Japan",
    capital: true, population: 9000000,
    regions: ["kanto", "honshu"] });
citiesRef.doc("BJ").set({
    name: "Beijing", state: null, country: "China",
    capital: true, population: 21500000,
    regions: ["jingjinji", "hebei"] });

Assuming the default automatic indexing settings, Cloud Firestore updates one ascending single-field index per non-array field, one descending single- field index per non-array field, and one array-contains single-field index for the array field. Each row in the following table represents an entry in a single-field index:

Collection Field indexed Query scope
cities name Collection
cities state Collection
cities country Collection
cities capital Collection
cities population Collection
cities name Collection
cities state Collection
cities country Collection
cities capital Collection
cities population Collection
cities array-contains regions Collection

Queries supported by single-field indexes

Using these automatically created single-field indexes, you can run simple queries like the following:

Web
const stateQuery = citiesRef.where("state", "==", "CA");
const populationQuery = citiesRef.where("population", "<", 100000);
const nameQuery = citiesRef.where("name", ">=", "San Francisco");

You can also create in and compound equality (==) queries:

Web
citiesRef.where('country', 'in', ["USA", "Japan", "China"])

// Compound equality queries
citiesRef.where("state", "==", "CO").where("name", "==", "Denver")
citiesRef.where("country", "==", "USA")
         .where("capital", "==", false)
         .where("state", "==", "CA")
         .where("population", "==", 860000)

If you need to run a compound query that uses a range comparison (<, <=, >, or >=) or if you need to sort by a different field, you must create a composite index for that query.

The array-contains index allows you to query the regions array field:

Web
citiesRef.where("regions", "array-contains", "west_coast")
// array-contains-any and array-contains use the same indexes
citiesRef.where("regions", "array-contains-any", ["west_coast", "east_coast"])

Queries supported by composite indexes

Cloud Firestore uses composite indexes to support compound queries not already supported by single-field indexes. For example, you would need a composite index for the following queries:

Web
citiesRef.where("country", "==", "USA").orderBy("population", "asc")
citiesRef.where("country", "==", "USA").where("population", "<", 3800000)
citiesRef.where("country", "==", "USA").where("population", ">", 690000)
// in and == clauses use the same index
citiesRef.where("country", "in", ["USA", "Japan", "China"])
         .where("population", ">", 690000)

These queries require the composite index below. Since the query uses an equality (== or in) for the country field, you can use an ascending or descending index mode for this field. By default, inequality clauses apply an ascending sort order based on the field in the inequality clause.

Collection Fields indexed Query scope
cities (or ) country, population Collection

To run the same queries but with a descending sort order, you need an additional composite index in the descending direction for population:

Web
citiesRef.where("country", "==", "USA").orderBy("population", "desc")

citiesRef.where("country", "==", "USA")
         .where("population", "<", 3800000)
         .orderBy("population", "desc")

citiesRef.where("country", "==", "USA")
         .where("population", ">", 690000)
         .orderBy("population", "desc")

citiesRef.where("country", "in", ["USA", "Japan", "China"])
         .where("population", ">", 690000)
         .orderBy("population", "desc")
Collection Fields indexed Query scope
cities country, population Collection
cities country, population Collection

To avoid performance loss caused by index merging, we recommend that you create a composite index to combine an array-contains or array-contains-any query with additional clauses:

Web
citiesRef.where("regions", "array-contains", "east_coast")
         .where("capital", "==", true)

// array-contains-any and array-contains use the same index
citiesRef.where("regions", "array-contains-any", ["west_coast", "east_coast"])
         .where("capital", "==", true)
Collection Fields indexed Query scope
cities array-contains tags, (or ) capital Collection

Queries supported by collection group indexes

To demonstrate an index with collection group scope, add a landmarks sub-collection to some of the city documents:

Web
var citiesRef = db.collection("cities");

citiesRef.doc("SF").collection("landmarks").doc().set({
    name: "Golden Gate Bridge",
    category : "bridge" });
citiesRef.doc("SF").collection("landmarks").doc().set({
    name: "Golden Gate Park",
    category : "park" });

citiesRef.doc("DC").collection("landmarks").doc().set({
    name: "National Gallery of Art",
    category : "museum" });
citiesRef.doc("DC").collection("landmarks").doc().set({
    name: "National Mall",
    category : "park" });

Using the following single-field index with collection scope, you can query a single city's landmarks collection based on the category field:

Collection Fields indexed Query scope
landmarks (or ) category Collection
Web
citiesRef.doc("SF").collection("landmarks").where("category", "==", "park")
citiesRef.doc("SF").collection("landmarks").where("category", "in", ["park", "museum"])

If you're interested in querying the landmarks across all cities, for example, you run this query on the collection group that consists of all landmarks collections. You must also enable a landmarks single-field index with collection group scope:

Collection Fields indexed Query scope
landmarks (or ) category Collection group

With this index enabled, you can query the landmarks collection group:

Web
var landmarksGroupRef = db.collectionGroup("landmarks");

landmarksGroupRef.where("category", "==", "park")
landmarksGroupRef.where("category", "in", ["park", "museum"])

To run a collection group query that returns filtered or ordered results, you must enable a corresponding single-field or composite index with collection group scope. Collection group queries that don't filter or order results, however, do not require any additional index definitions.

For example, you can run the following collection group query without enabling an additional index:

Web
db.collectionGroup("landmarks").get()

Index entries

Your project's configured indexes and the structure of a document determine the number of index entries for a document. Index entries count towards the index entry count limit.

The following example demonstrates the index entries of a document.

Document

/cities/SF

city_name : "San Francisco"
temperatures : {summer: 67, winter: 55}
neighborhoods : ["Mission", "Downtown", "Marina"]

Single-Field indexes

  • city_name ASC
  • city_name DESC
  • temperatures.summer ASC
  • temperatures.summer DESC
  • temperatures.winter ASC
  • temperatures.winter DESC
  • neighborhoods Array Contains (ASC and DESC)

Composite indexes

  • city_name ASC, neighborhoods ARRAY
  • city_name DESC, neighborhoods ARRAY

Index entries

This indexing configuration results in the following index entries for the document:

Index Indexed data
Single-field index entries
city_name ASC city_name: "San Francisco"
city_name DESC city_name: "San Francisco"
temperatures.summer ASC temperatures.summer: 67
temperatures.summer DESC temperatures.summer: 67
temperatures.winter ASC temperatures.winter: 55
temperatures.winter DESC temperatures.winter: 55
neighborhoods Array Contains ASC neighborhoods: "Mission"
neighborhoods Array Contains DESC neighborhoods: "Mission"
neighborhoods Array Contains ASC neighborhoods: "Downtown"
neighborhoods Array Contains DESC neighborhoods: "Downtown"
neighborhoods Array Contains ASC neighborhoods: "Marina"
neighborhoods Array Contains DESC neighborhoods: "Marina"
Composite index entries
city_name ASC, neighborhoods ARRAY city_name: "San Francisco", neighborhoods: "Mission"
city_name ASC, neighborhoods ARRAY city_name: "San Francisco", neighborhoods: "Downtown"
city_name ASC, neighborhoods ARRAY city_name: "San Francisco", neighborhoods: "Marina"
city_name DESC, neighborhoods ARRAY city_name: "San Francisco", neighborhoods: "Mission"
city_name DESC, neighborhoods ARRAY city_name: "San Francisco", neighborhoods: "Downtown"
city_name DESC, neighborhoods ARRAY city_name: "San Francisco", neighborhoods: "Marina"

Indexes and pricing

Indexes contribute to the storage costs of your application. For more information about how to calculate storage size for indexes, see Index entry size.

Use index merging

Although Cloud Firestore uses an index for every query, it doesn't necessarily require one index per query. For queries with multiple equality (==) clauses and, optionally, an orderBy clause, Cloud Firestore can re-use existing indexes. Cloud Firestore can merge the indexes for simple equality filters to build the composite indexes needed for larger equality queries.

You can reduce indexing costs by identifying situations where you can use index merging. For example, in a restaurants collection for a restaurant rating app:

  • restaurants

    • burgerthyme

      name : "Burger Thyme"
      category : "burgers"
      city : "San Francisco"
      editors_pick : true
      star_rating : 4

This app uses queries like the following. The app uses combinations of equality clauses for category, city, and editors_pick while always sorting by ascending star_rating:

Web
db.collection("restaurants").where("category", "==", "burgers")
                            .orderBy("star_rating")

db.collection("restaurants").where("city", "==", "San Francisco")
                            .orderBy("star_rating")

db.collection("restaurants").where("category", "==", "burgers")
                            .where("city", "==", "San Francisco")
                            .orderBy("star_rating")

db.collection("restaurants").where("category", "==", "burgers")
                            .where("city", "==" "San Francisco")
                            .where("editors_pick", "==", true )
                            .orderBy("star_rating")

You could create an index for each query:

Collection Fields indexed Query scope
restaurants category, star_rating Collection
restaurants city, star_rating Collection
restaurants category, city, star_rating Collection
restaurants category, city, editors_pick, star_rating Collection

As a better solution, you can reduce the number of indexes by taking advantage of Cloud Firestore's ability to merge indexes for equality clauses:

Collection Fields indexed Query scope
restaurants category, star_rating Collection
restaurants city, star_rating Collection
restaurants editors_pick, star_rating Collection

Not only is this set of indexes smaller, it also supports an additional query:

Web
db.collection("restaurants").where("editors_pick", "==", true)
                            .orderBy("star_rating")

Indexing limits

The following limits apply to indexes. For more information about quotas and limits, see Quotas and Limits.

Limit Details
Maximum number of composite indexes for a database
Maximum number of single-field configurations for a database

One field level configuration can contain multiple configurations for the same field. For example, a single-field indexing exemption and a TTL policy on the same field count as one field configuration towards the limit.

Maximum number of index entries for each document

40,000

The number of index entries is the sum of the following for a document:

  • The number of single-field index entries
  • The number of composite index entries

To see how Cloud Firestore turns a document and a set of indexes into index entries, see this index entry count example.

Maximum number of fields in a composite index 100
Maximum size of an index entry

7.5 KiB

To see how Cloud Firestore calculates index entry size, see index entry size.

Maximum sum of the sizes of a document's index entries

8 MiB

The total size is the sum of the following for a document:

  • The sum of the size of a document's single-field index entries
  • The sum of the size of a document's composite index entries
  • Maximum size of an indexed field value

    1500 bytes

    Field values over 1500 bytes are truncated. Queries involving truncated field values may return inconsistent results.

    Indexing best practices

    For most apps, you can rely on automatic indexing and the error message links to manage your indexes. However, you may want to add single-field exemptions in the following cases:

    Case Description
    Large string fields

    If you have a string field that often holds long string values that you don't use for querying, you can cut storage costs by exempting the field from indexing.

    High write rates to a collection containing documents with sequential values

    If you index a field that increases or decreases sequentially between documents in a collection, like a timestamp, then the maximum write rate to the collection is 500 writes per second. If you don't query based on the field with sequential values, you can exempt the field from indexing to bypass this limit.

    In an IoT use case with a high write rate, for example, a collection containing documents with a timestamp field might approach the 500 writes per second limit.

    TTL fields

    If you use TTL (time-to-live) policies, note that the TTL field must be a timestamp. Indexing on TTL fields is enabled by default and can affect performance at higher traffic rates. As a best practice, add single-field exemptions for your TTL fields.

    Large array or map fields

    Large array or map fields can approach the limit of 40,000 index entries per document. If you are not querying based on a large array or map field, you should exempt it from indexing.

    If you are using queries with range and inequality operators on multiple fields, see the indexing considerations that you should consider to optimize the performance and cost of Cloud Firestore queries

    For more information about how to resolve indexing issues (index fanout, INVALID_ARGUMENT errors) see the troubleshooting page.