This page introduces Atlas Data Federation key concepts, detailing how virtual databases, collections, and storage configurations enable unified data access and querying across multiple sources.
Concept | Description |
|---|---|
Data Federation | Data Federation is a strategy that separates compute from storage. When you use Data Federation, you associate data from multiple physical sources into a single virtual source of data for your applications. This enables you to query your data from a single endpoint without physically copying or moving it. |
Federated Database Instance | A federated database instance is a deployment of Atlas Data Federation. Each federated database instance contains virtual databases and collections that map to data in your data stores. |
Virtual Database | A virtual database is a grouping of collections that organizes your data into a single virtual location. |
Virtual Collection | A virtual collection is a grouping of MongoDB documents, similar to a table in a relational database, within a single virtual database. Collections don't require a fixed schema. |
Data Store | A data store refers to the location of your data. Atlas Data Federation supports the following data stores:
A federated database instance is limited to one instance of AWS S3, Azure Blob Storage, and Google Cloud Storage at the same time. Similarly, Data Federation across Atlas projects is unavailable. To learn more, see Atlas Data Federation Limitations. |
Storage Configuration | Storage configuration contains mappings between your virtual databases, collections, and data stores. You can define these mappings to access and run queries against your data. |