Create Schema
On this page
Create Schema Using the Atlas UI
Schema generation happens automatically when you create an Atlas SQL quick start connection.
In Atlas, go to your federated database instance for your project.
If it's not already displayed, select the organization that contains your project from the Organizations menu in the navigation bar.
If it's not already displayed, select your project from the Projects menu in the navigation bar.
In the sidebar, click Data Federation under the Services heading.
The Data Federation page displays.
Create Schema Using mongosh
The sqlGenerateSchema command generates an Atlas SQL schema for the
specified collections or views.
Syntax
db.runCommand({ sqlGenerateSchema: 1, sampleNamespaces: [<namespace>], sampleSize: <int>, setSchemas: true|false })
Parameters
Parameter | Type | Description | Necessity |
|---|---|---|---|
| array of strings | Specifies the comma-separated list of namespaces for which to generate
schemas. A namespace
includes the database name, a dot (.) separator, and the collection or
view name (i.e. | Optional |
| integer | Specifies the number of documents to use as a sample to create the schema.
If omitted, defaults to | Optional |
| boolean | Specifies whether or not to store the generated schema for the collection or view. Value can be one of the following:
If omitted, defaults to | Optional |
Output
The command returns the following if the command succeeds.
{ "ok" : 1, "schemas" : [ { "databaseName" : "<database-name>", "namespaces" : [ { "name" : "<collection-name>", "schema" : { "version" : NumberLong(1), "jsonSchema" : {} } } ] }, ... ] }
The schemas object contains the following fields.
Parameter | Type | Description |
|---|---|---|
| string | Name of the database. |
| array of objects | Name and generated schema of each collection or view. |
| string | Name of the collection or view. |
| document | Schema of the collection or view. |
| integer | Format version of the schema. Value is always 1. |
| document | JSON schema of the collection or view. The JSON schema can contain the following fields:
To learn more about these fields, see JSON Schema Keywords. |
If you set the schema for the collection or view using the
setSchemas option, you can verify that the command
succeeded by running the sqlGetSchema command. The
sqlGetSchema command metadata.description should
contain the following value:
"set using sqlGenerateSchema with setSchemas = true"
Examples
Consider a collection named egData in a database named
sampleDB with the following documents:
{"a": {"b": {"c": [1, 2, 3]}}, "s": 1} {"a": {"b": {"c": [4, 5, 6]}}, "s": 2} {"a": {"b": [7, 8, 9]}, "s": 3} {"a": {"b": {"c": []}}, "s": 4} {"a": {"b": {"c": "hello"}}, "s": 5} {"a": {"b": {"c": {"d": 1}}}, "s": 6} {"a": {"b": {"c": null}}} {"s": 7}
The examples below use the sqlGenerateSchema command to
generate a schema for the above collection.
Basic Example
The following command generates a schema for the collection
named sampleDB.egData in the storage configuration. The
command uses two randomly selected documents from the
collection to create the schema because the sampleSize
is 2. The command does not set the schema for the
collection because the setSchemas option is not specified
with the command and defaults to false.
db.runCommand({ sqlGenerateSchema: 1, sampleNamespaces: ["sampleDB.egData"], sampleSize: 2 })
The previous command returns the following output. To learn more about the fields in the output, see Output.
{ "ok" : 1, "schemas" : [ { "databaseName" : "sampleDB", "namespaces" : [ { "name" : "egData", "schema" : { "version" : NumberLong(1), "jsonSchema" : { "bsonType" : [ "object" ], "properties" : { "a" : { "bsonType" : [ "object" ], "properties" : { "b" : { "bsonType" : [ "object" ], "properties" : { "c" : { "bsonType" : [ "array" ], "items" : { "bsonType" : [ "int" ] } } } } } }, "s" : { "bsonType" : [ "int" ] } } } } } ] } ] }
Generate and Set Schema Example
The following command generates a schema for the collection
named sampleDB.egData in the storage configuration. The command
uses up to 1000 documents in the collection to create the schema
because the sampleSize option is not specified with the command
and defaults to 1000. The command sets the generated schema as
the schema to use for the collection because the setSchemas
option is set to true.
db.runCommand({ sqlGenerateSchema: 1, sampleNamespaces: ["sampleDB.egData"], setSchemas: true })
The previous command returns the following output. To learn more about the fields in the output, see Output.
{ "ok" : 1, "schemas" : [ { "databaseName" : "sampleDB", "namespaces" : [ { "name" : "egData", "schema" : { "version" : NumberLong(1), "jsonSchema" : { "bsonType" : [ "object" ], "properties" : { "a" : { "bsonType" : [ "object" ], "properties" : { "b" : { "bsonType" : [ "object", "array" ], "properties" : { "c" : { "bsonType" : [ "array", "string", "object", "null" ], "properties" : { "d" : { "bsonType" : [ "int" ] } }, "items" : { "bsonType" : [ "int" ] } } }, "items" : { "bsonType" : [ "int" ] } } } }, "s" : { "bsonType" : [ "int", "object" ] } } } } ] } ] }
Errors
The command returns the following error if the command fails:
"failedNamespaces": [ { "namespace" : "<db.ns>", "error" : "no documents found in sample namespace" } ]
The above error is returned if the specified namespaces do not exist in the storage configuration or are empty. This error is also returned if the schema could not be set for a given namespace.