After failing over operations in the event of a disaster, you'll eventually need to fail back to either the original data center or to a new one if the original is beyond repair. If you chose not to fail over operations, you'll need to restore the data from tape or another point-in-time copy in order to get back to work.
While it may seem like the worst is now behind you, it's important to approach the restoration process with as much care as the disaster. Proper planning for restoration is as vital as the preparations you make for the actual failure.
If you're working from a remote data center, remember that restoring data to the primary data center may very well be an independent process from the fail-back operations themselves. Fail-back operations can typically take as little time as the failover took, but restoring the data could take hours or even days, depending on how much data you need to move and the kind of networking you're using to move it over.
To help ensure faster data restoration, consider planning for burst-able pipes or leased lines. While this option will add overhead to your day-to-day budget, it can dramatically reduce the amount of time it will take to restore normal operations.
As for the fail-back itself, this is an event you can control when it comes to time. Take advantage of that fact, and make sure to schedule a well-planned maintenance window.
During this maintenance window, you can bring the original systems back up and test them appropriately. Then, only when you're positive you've resolved the issues that caused the failure, reroute your end users back to their home systems.
Restoring from tape means you'll be working with more limitations. You have a finite amount of time (based on your recovery time objectives) to restore the data back to the systems in question, and you have to do it as soon as possible. That means planning a maintenance window is unfortunately out of the question.
Obtain your tapes, and move them to the site of the servers you'll be sending the data back to. These servers could either be the normal servers (after you've corrected minor hardware or software issues), or they could be entirely new servers.
If you're restoring data to the same servers, make sure you've completely tested them before beginning the restoration process. Making sure you haven't missed a secondary issue can help you avoid last-minute problems when bringing systems back online.
Expect to contend with potentially lost backup tapes, faulty backup systems, and other unanticipated annoyances. Minor issues like these won't necessarily torpedo your restoration efforts, but they could easily add downtime to the overall process.
If you're restoring to new hardware, expect to encounter these issues as well as a few more since you're working with new equipment. Most notably, you may not have the proper tape hardware to perform the restoration, which could take quite some time to replace. While this isn't an insurmountable problem, it definitely could put a crimp in your time budget—not to mention, your fiscal budget.
Planning for data restoration isn't something many people think about, but it could become a critical chink in your disaster recovery armor. Including a thorough plan for restoring systems in your organization's business continuity plan can make the real difference when recovering from a disaster.
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.