How do I model relationships and schema changes?

Definition:

Data modeling describes how records relate and how the application retrieves and changes them. Schema evolution is the process of changing that structure while old data and possibly old application versions still exist. MongoDB's flexible document shape makes gradual migration possible, but compatibility still needs planning.

Key points:

• One-to-one: embed or store an ID; one-to-many: decide based on size and access.

• Many-to-many: usually use references or a separate linking collection.

• For schema changes, support old documents while gradually updating them.

• Consider a version field, tolerant reads and a background backfill when introducing a required field to existing documents.

Interview point:

• Explain how an old document will be read during deployment, not only how a new document will be written.