What is normalization, and what are primary and foreign keys?

Question

Accepted Answer

Normalization is organizing data into tables so each fact lives in exactly 1 place, which removes redundancy and the update anomalies that come with it. If a customer address is stored once, you update it once, rather than chasing 5 copies scattered across rows that drift out of sync. The keys are what hold a normalized schema together. A primary key is the column, or combination of columns, that uniquely identifies each row in a table. It enforces uniqueness and gives other tables a stable handle to point at. A foreign key is a column in one table that references another table's primary key, and it enforces referential integrity, so the database will not let you reference a row that does not exist, and you choose what happens on delete, cascade, restrict, or set null. The nuance I would add is that normalization is the default for transactional data because it keeps data correct, but denormalization is a deliberate and legitimate trade-off. Sometimes I duplicate a field to avoid a 4 table join on a hot read path, accepting the redundancy and taking on the job of keeping the copies in sync. So I normalize for integrity, and I denormalize consciously for read performance when the access pattern really justifies it.

Answer

When would you deliberately denormalize?

For read-heavy access patterns where joins are too expensive, you duplicate data (or precompute a read model) to serve reads fast, accepting redundancy and the cost of keeping copies in sync. It is a performance trade-off, not a mistake.

Answer

What does a foreign key constraint protect against?

Orphaned/invalid references: you cannot insert a row referencing a non-existent parent, and deletes are controlled (cascade/restrict/set null). It enforces referential integrity at the database level.

What is normalization, and what are primary and foreign keys?

What is normalization, and what are primary and foreign keys?

The follow-ups they ask next

What the interviewer is listening for

What sinks the answer

If you genuinely do not know

Keep going with databases

Knowing the answer is not the same as recalling it under pressure