Data Modeling and Migrations

Bad schema decisions are expensive to fix later. Plan for change from the start.

Data modeling essentials

• Normalize where needed, denormalize where useful
• Define clear ownership of data entities
• Use indexes intentionally
• Track constraints and invariants in schema

Migration safety

• Make migrations forward-compatible
• Avoid long locks on large tables
• Backfill data in batches
• Validate before and after migration
• Keep rollback strategy ready

Resources

• Database Normalization
• PostgreSQL ALTER TABLE
• Prisma Migrate
• Liquibase Docs