When designing sufficiently large relational databases, a decision is often made to deviate from normal form - “denormalization”. In such cases, triggers can help ensure data integrity.

Read the article on Habr