Data Modeling in a Non-Relational World
summary
Relational schema rules break when using documents; model around aggregates and coherent independent isolated units instead of foreign keys. RavenDB supports embedding for value objects, referencing for independent entities, and an include/load pattern to avoid extra round trips. Practical numbers: a profiler found spikes up to 17,000 queries per pageview while typical problematic apps hit dozens; aim for tens not thousands. Watch for obese documents (megabyte-sized) and duplicate temporal data mistakes (e.g., address or payroll) that corrupt history.
AI-written summary · about
chapters (17)
- 0:00Introduction to talk
- 1:16Relational modeling history
- 6:48Wikipedia caching example
- 8:36Order page performance issues
- 13:34Schema constraints in modeling
- 17:29JSON document modeling basics
- 19:17Coherent independent isolated documents
- 26:48Foreign keys vs document references
- 29:22Data duplication trade-offs
- 33:39Temporal data challenges
- 35:07Embedding vs referencing data
- 36:54Obese vs electric documents
- 39:03One-to-many associations
- 41:12Lazy loading pitfalls
- 46:40Many-to-many associations
- 50:04Query-time data shaping
- 53:29Q&A session
related talks
44mThe Developer's Guide to Data Modelling with Document Databases
46mFrom Tables to Documents -- Changing Your Database Mindset
26mNoSQL Data Modeling for Front-end Developers
42mChoosing the right database
48mJSON Data Modeling
59mData Modeling for
19mRelational or Non-Relational? Choosing the Best Database for the Task
36mMongoDB Schema Design - best practices