What are the trade-offs between normalized and denormalized data structures in a modern data warehouse environment?
