I was asked to answer a question in a phone interview for Amazon.
The question is as follow:
Image there is a website generates user access log in Terabytes scale every day. Each log contains many different information such as username, visiting time, location, page visited, and etc.
what's your solution if we want to only store certain info(for example, user, login time, location) to a sql database. How to optimize the process.

This is an open question. So I believe there is no correct answer. what is your thought?