Nowadays computer components are cheap, and in combination with high-speed Internet they offer a whole new range of possibilities like centralized information storage.
Governments and commercial parties are massively adopting this principle, and a few scenarios are not always carefully considered.
The advantages are clear and include lower Maintenance costs and easier access to information.
Centralized information systems, especially the important ones are always built redundant and distributed. The hardware used in the servers is mostly identical for every machine, and the operating systems are copied from the same image, with the same version of the same software. The result is that it's much easier to make critical and self-reproducing mistakes.
Every large information system has allot of users (and super users), accounts are often compromised, only in the case of the Dutch police force for instance, all the criminal records stored in that database, (lets say the whole of the Netherlands for this example) are accessible with the right permissions. But thats not the only problem, you may have heard of bot-nets before from TV (a large group of networked machines performing the same task). All these botnet bots run the same software version, and could be compromising (taking over) a whole cluster of servers by making use of a massively implemented security hole as described by 'The human factor'. Say hello to anarchy..
Scenario 'Impending doom' (total control over a influential system) is becoming more and more a realistic possibility. Before you store information in one system, an administrator should think, am I making a smart decision?
Jonathan