Denormalization

Denormalization of data is another important concept, especially when it comes to querying efficiency.

In an enterprise, everything is about balance. Most organizations that deal with data do not completely adhere to the complex form of normalization. The reason for this is simple: normalization may kill efficiency when you need to query a large amount of data. Therefore, while we gain in terms of data integrity when normalizing data, we pay a lot in terms of processing power if we ever have to query data.

For instance, consider the table in Figure 3.17. If a user has two email addresses, would you have a single column and add both email addresses to it? You could, but that would be non-compliant with the first normal form of data. How do you work around this issue? Create a new table with an employee ID and email address? But then, the employee would have two email addresses and cannot be uniquely identified. Therefore, you would need to add a primary key to...