The EAV pattern stores data as rows: `[EntityID, Attribute, Value]`. When is this acceptable?

1. A When the collection of target attributes is vast, sparse, and highly dynamic (e.g., Medical Records) where defining hundreds of NULL columns would be inefficient.
2. B When modeling standard user authentication profiles where the attributes are rigid, predictable, and shared by every single member of the application.
3. C When implementing high-frequency financial ledgers that require strict ACID compliance and absolute mathematical consistency for every multi-row transaction.
4. D When designing high-performance caching layers that prioritize O(1) read latency for simple key-based lookups over schema flexibility.