A good data model improves application performance, developer productivity, and reduces overall development costs.
NoSQL document databases like MongoDB use collections of BSON documents with flexible schemas, unlike the rigid tables of SQL.
One-to-many relationships can be modeled by either embedding related data within a single document or by referencing it using IDs.
Use embedding for one-to-few relationships and referencing for one-to-zillions, while considering if entities are separate logical units for one-to-many cases.
Combine referencing with embedding of frequently accessed fields to reduce queries while avoiding overly large documents and data duplication.
Accept duplication for historical data and manage necessary updates for stale data using batch operations triggered by change streams.
Effective data modeling is driven by application use cases, requiring identification of logical units, relationships, and critical queries.
Sensory-Minds GmbH
Offenbach am Main, Germany
Hubert Burda Media
München, Germany
fulfillmenttools
Köln, Germany
44:51Stanimira Vlaeva
47:16Maxime Beugnet
31:08Hartmut Armbruster
51:04Noam Honig
32:56Laurie Voss
27:27Michele Riva
24:22Jürgen Pilz
51:10Laurie Voss
Jobs that call for the skills explored in this talk.
Antal International
Barcelona, Spain
Sept Lieues
Paris, France
Da Vinci Engineering GmbH
Stuttgart, Germany
Antal International
Municipality of Valencia, Spain
Percona