Home/ Questions/Q 8200463
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-06-07T06:27:30+00:00

Lets say I have an Illustration entity which is an aggregate root. That entity

  • 0

Lets say I have an Illustration entity which is an aggregate root. That entity contains some data about the artwork and is persisted in a SQL database while the artwork itself is persisted on Amazon S3. Also, I would want to save some scaled-down or thumbnail versions of the artwork so I introduced a Blob entity in many-to-one relationship with Illustration for representing the binary data of the artwork in various versions.

Now I wonder how should i design the persistence of Blobs. Amazon S3 is a kind of database (please don’t start the flame what is a true database here 😉 ), just different than SQL and I think that it should be abstracted like so, that means by a Repository. So I would have a BlobRepository where I could store artwork data. On the other hand – in this domain Blob is definitely not an aggregate root – it is always used as a part of Illustration aggregate. So it should’t have its own repository.

So maybe Amazon S3 should not be treated as a persistence technology, but rather as a generic external service, next to EmailSender, CurrencyConverter etc.? If so, where should I inject this service? Into Illustration entity methods, IllustrationsRepository, application service layer?

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    First of all when dealing with DDD there is no many-to-one or any RDBMS concepts, because in DDD the db does not exist, everyhthing is sent to a Repository. If you’re using an ORM, know that the ORM entities are NOT Domain entities, they are Persistence objects.

    That being said, I think the Illustration Repository should abstract both the RDBMS and S3. These are persistence implementation details, the repo should deal with them. Basically the repo will receive the Illustration AR, which will be saved partly in RDBMS and partly as a blob in S3.

    So the Domain doesn’t know and shouldn’t know about Amazon S3, maybe tomorrow you’ll want to use Azure Db, why should the Domain care about it?. It’s the Repository’s responsibility to deal with it.

    • 0