Comparing Amazon DocumentDB and MongoDB

Document databases store data in JSON-like documents, which makes them inherently easy for developers to work with. MongoDB is a modern document database that contains many of the powerful features found in popular relational databases. These include ACID-compliant transactions, schema governance, enterprise-grade security, and more.

MongoDB offers the Atlas Data Federation engine, which allows users to quickly and easily query data in any format on Amazon S3 using the MongoDB Query API.

DocumentDB does not contain any MongoDB code, supports a limited subset of MongoDB features, and is more costly and difficult to manage.

DocumentDB secondary indexes are built to behave similarly to their MongoDB counterparts. However, DocumentDB's full-text search and vector search indexes are more limited in quantity and functionality.

Both DocumentDB and MongoDB support automated backups, as well as incremental and on-demand snapshots. However, DocumentDB requires the creation of a new cluster for restoring a backup. MongoDB simplifies this process by enabling backup restoration to existing clusters.

For non-sharded clusters, DocumentDB offers billing options similar to Atlas: instance-based billing and I/O-optimized billing. However, global clusters only offer an I/O-based pricing model, and DocumentDB's version of sharded clusters (elastic clusters) is billed using a more abstract, difficult-to-predict vCPU-consumption model.