Abstract
Disaster Recovery Plans is a crucial part of the life of an IT center in an organization; it contains policies and procedures to be applied before, during, and after a disaster of an IT system. However, an important part of the disaster recovery process comes when the disaster occurs up until the disaster recovery plan is activated. This is precious time to detect and declare a disaster especially in critical systems. This is important to control the MTD (Mean Tolerable Downtime). Here, we present a distributed disaster detection system that is based on agents. Agents are to be located on different servers in a data center and the will communicate with a central management unit. The DDDS is aimed to complement the work of an existing disaster recovery system, or to run stand-alone if no disaster recovery system exists and act as a warning tool aiding the system administrator.