I have to agree that the concept of disaster recovery should include the entire system infrastructure, including total site loss. But as mentioned earlier, if you don’t even have valid backups for the database, recovery may be impossible.

And so Steve’s title is accurate for a large part of the internet focused, and micro ~ PC based systems. Disaster Recovery is quite often an afterthought.

Mainframe shops almost always have DR plans that are tested. PC based shops rarely have any DR plans, and often don’t even test the database backups. Hardware fails. So plan accordingly. I once was in a meeting with an overly confident JD Edwards consultant going on how the system was solid: “…it is a 400.” (reference to an IBM A/S 400, normally a really solid piece of hardware. ) Less than a week later, the A/S 400 had a hardware failure, and JD Edwards was offline for about 4 days. So much for that consultant's bravado!