An Internet retailer's database was recently exploited by a foreign criminal organization via a remote attack. The initial exploit resulted in immediate root-level access. What could have been done to prevent this level of access being given to the intruder upon successful exploitation?