When and how does the database slow down the site?

Question

I read O'Reilly's book, which says that 70-80% of site performance can be optimized through the interface. This may mean that database queries may account for less than 20 or 30% of the site’s performance. However, I have seen huge sites such as Facebook and Twitter, devoting so much time to optimizing queries (via query caching, normalization, etc.). When a database becomes so critical to site performance that it can account for more than the aforementioned percentage? Oh, and by performance, I mean in the context of speed, download speed in particular.

Answer 1

When does a database become so critical to site performance that it can make up more than the aforementioned percentage?

Once you have optimized the front-end to max, the relative influence of the background content will become much greater.

In addition, very large sites face scalability issues. They may be able to serve one user very quickly, but as the load grows, it becomes slower for everyone. Part of the client part is distributed among users' machines, so it scales very well. But the basic part is shared by all and will become a bottleneck.

Answer 2

, . - : - , - , . , - . , ( ) , / .