Can you give a situation where abandoning database normalization is justified?

Opting out of database normalization can be justified in cases where high read performance and minimizing the number of table joins are important. For example, in systems with a high volume of read requests and relatively infrequent data updates, denormalization helps speed up data access by keeping related data in a single table.

Example: In analytical or reporting systems, where data rarely changes but is read frequently, storing aggregated or redundant data can accelerate queries.

However, it is important to consider potential issues with data integrity and the complexity of updates.