Five Tips for a Sound Disaster Recovery Plan
Organizations of all sizes are dealing with more data today than ever before. Much of this data is so integrated into business processes that it ensures businesses keep functioning. Our data dependency makes it vital to set up systems that protect critical data on a day-to-day basis, so now matter what—power outage, device theft or human error—business doesn’t skip a beat.
However, it is also important to build a robust Disaster Recovery Plan for your data in the event of a worst-case scenario. Developing an entire Disaster Recovery Plan from scratch can be an overwhelming process, so I have come up with five simple tips to help you get started in building your own sound and foolproof plan:
1. Assume the Worst
Let’s start off with a bit of humility: there is no such thing as a perfect plan. We all would like it if everything always went as planned, but with a Disaster Recovery Plan you cannot afford any luxuries. You must assume that each step of your Disaster Recovery Plan could fail, and you have to prepare for “mini-Disasters” and guard against Murphy’s Law. For each step of your Disaster Recovery Plan, you should ideally have a “Plan B” and even a “Plan C.” Of course, in reality, many SMBs do not have the time and resources to build infinite redundancy into their Disaster Recovery Plans. Therefore to make the best possible decisions in evaluating which steps require alternatives, you will need to determine which steps could pose the most risk.
2. More Risk, More Redundancy
The person who is making your Disaster Recovery Plan, be that an administrator or IT expert, needs to dedicate a portion of their time to identify which steps are the most complex and thus carry the most risk. For those risky steps, it is essential to have a Plan B -- and never a bad idea to have a Plan C as well. For example, let’s say you are planning to restore a server from your most recent tape backup in the event of a disaster. If that fails, you would need to restore from a second set of tapes. Finally, if that option goes kerplunk, you could restore it from an alternative backup located in Cloud Storage. If you haven’t thought ahead, you could lose valuable time and even worse invaluable data.
3. Choose Simple and Flexible Backup Solutions
Now, when it comes time to choose your backup solution, it is important to pay attention to a few qualities. One is simplicity. It may seem that the most important thing is that your systems are safe and backed up, but remember - it can be a huge pain or worse — impossible — to restore them if the backup software is complex. For example, File-Based Backup Software may require you to reinstall your entire operating system and all your applications from scratch. A Simpler Disk Imaging Solution will allow you to restore everything in a single pass – so everything is exactly the way it was before the disaster. Rule of thumb – keep it simple stupid and image software is a much simpler and safer bet. Additionally, look for a flexible option that will give you the ability to choose alternative routes when it comes to hardware or even a platform. Hardware changes rapidly nowadays and it is important that you do not lock-in your data and your systems. It then becomes difficult to restore to a new machine, whether that be Physical from a different vendor or even Virtual.
4. Be Critical
It is easy to be critical of others’ work, but in the case of Backup and Disaster Recovery, you must also look at your own efforts with a clear and unbiased eye. Being critical of yourself and your system can help you spot errors and weaknesses before a disaster brings it all crashing down. Also, don’t be an island—you have friends. Ask for constructive feedback from your team, your superiors, your vendors, your System Integrator or anyone else who might be able to help you spot problems before it is too late. A good plan is a resilient one, so do not be afraid to poke holes in your own handiwork – and let it be ridicules by others.
5. Test, Test and Test Again
Practice makes perfect, they say, and when it comes to Disaster Recovery the key is to test your plan repeatedly. A Disaster Recovery Plan will not be a smooth or dependable, unless you have put it to the test and effectively implemented it. Before that moment your ability to recover is all based on theory.
You should test it on actual Disaster Recovery hardware, as well as on virtual machines. Another alternative is to split up into teams and task one team with “sabotaging” the other’s Disaster Recovery environment. Then let the other team attempt to recover their data and find out if your plan works. This way you can identify issues, resolve them and improve your Disaster Recovery Plan. Remember, one test is never enough, and keep working to improve your plan over time. Like a fire drill, it is best if you never have to actually use it, but if you do, you will be prepared.
What tips do you have for other SMBs looking to improve their plans for disasters? Have you ever faced a major data disaster of your own?