While you have a variety of options for your staff's actions during a disaster, the different methods generally revolve around the same basic principles. In the case of any disaster, there is a critical point in time when you must act quickly to initiate your disaster response. This may sound like a simple, and obvious, part of the job, but keep in mind that many emergencies can cut you off from normal means of communication—and probably cut you off from the production data center as well.
The basic idea is to prepare for both emergencies that you can deal with locally and ones you'll need to handle remotely. Relatively speaking, local emergencies are typically the easiest to deal with. In these situations, you can command and control the failover operations from your primary location, especially if only one or a few critical systems fail.
If you have local failover systems in place, you can initiate these procedures at the servers themselves. If you don't have local failover systems, you can enact tape-restore procedures quickly and effectively from the same location from which the failure occurred, provided the server systems themselves are still operational.
For emergencies that involve the loss of the primary data center, or if you just can't get to the data center quickly, the first response becomes more complicated—but no less important.
You can, of course, set up automated failover systems that require no user interaction. However, this approach can be risky across WAN connections; a dropped data-link could accidentally fail over your organization.
Instead, it's a better idea to set up remote-access systems, such as the services provided by Microsoft in its Windows Server products or one of the many products designed for any number of operating systems. This permits you to digitally access the failover data center, even if you can't physically access the actual location.
In either case, it's vital that your staff is ready to go well before a disaster strikes. For critical data systems, you may have only a few minutes to get the systems back up and online.
However, you often don't have sufficient time to properly restore data from offline storage. So if tape is your only recovery system, this particular portion of failover operations wouldn't apply to your organization.
But if you've put replication and failover tools into place, you must ensure that you've cross-trained staff on how to fail over multiple critical systems. It's not an uncommon occurrence in a disaster situation for the person who has primary control over any given system to be unavailable.
After you've trained your staff, it's important to put a communication system into place that will quickly and efficiently alert all necessary staff to the fact that you're declaring an emergency. Keep in mind that e-mail may not be available, so relying solely on an e-mail alert won't cut it. Telephone lists, pager systems, "walkie-talkie"-enabled mobile devices, and other immediate communication systems offer options for alerting staff quickly when another method of communication fails.
How well can your organization deal with an emergency? Automatically sign up for our free Disaster Recovery newsletter, delivered each Tuesday, and make sure you're prepared for the next catastrophe.