The Modern Database Debate: Documents, Tables, or Documents in Tables?
summary
Compares PostgreSQL and MongoDB with concrete examples: a Spring Data demo app shows many SQL round trips for normalized schemas versus single-document reads in MongoDB, and MongoDB supports multi-document ACID transactions now; warns that putting everything as JSON in Postgres works for storage but PostgreSQL JSON indexing limits make many queries inefficient, whereas MongoDB supports compound indexes on nested/array fields and lets aggregation pipelines run server-side (watch the 16MB BSON aggregation result limit). Recommends modeling by access patterns and testing indexes, not labels.
AI-written summary · about
chapters (22)
- 0:00Introduction to db debate
- 4:50Database popularity trends
- 8:01Debunking database myths
- 11:26Modeling and complexity
- 14:08Data access and transactions
- 16:40Documents vs normalized data
- 20:06Data modeling strategies
- 23:02Modern design and DDD
- 25:08Duplication vs normalization
- 28:42Domain access patterns
- 30:21Spring Data example
- 31:34Query generation comparison
- 32:44Application-centric modeling
- 36:07Query performance tuning
- 39:17JSON in SQL databases
- 42:35Indexing complex data
- 48:01Choosing the right model
- 51:16Compound indexing insights
- 52:27Final database recommendations
- 53:30Best practices
- 54:31Handling references
- 57:20Database internals and wrap up
related talks
46mFrom Tables to Documents -- Changing Your Database Mindset
41mFrom Tables to Documents - Changing Your Database Mindset
38mFrom Tables to Documents—Changing Your Database Mindset
42mChoosing the right database
53mPostgres vs Mongo
53mNoSQL Best Practices for PostgreSQL
01h 00mFrom Relational Databases to NoSQL
26mDesigning for Document Databases in PostgreSQL