Over the years I have written a number of IT disaster recovery plans. The first, written for a mainframe manufacturer in the late 80s, looked more like a thesis when printed out and tried, probably unsuccessfully, to document and direct every possible eventuality. I doubt whether anyone else properly read it – it was certainly never tested.
Most small business customers that I come across don’t have an IT disaster recovery plan though quite a number agree that they should ‘think about it’. I suspect that this gap between intention and action is because many people’s image of a disaster recovery plan is of a large report that looks like a thesis when printed out.
Let’s think for a minute what we’re trying to achieve. The first task is to understand the structure of the technology in use. For even quite large firms we can usually do this diagrammatically on one side of A4. Using this as our template, we can then introduce breaks in to the structure and as questions like ‘What would happen if this didn’t work?’.
Armed with the answer to that question and perhaps half a dozen like it, you can annotate your diagram with the impact of these failure points. Your technology support person can then take the diagram away and answer a final set of questions for each failure point.
These questions are the same for any failure point.
1. Is it in our power to fix it?
2. If yes, how long will it take to fix it?
3. Is there something we could do or buy to shorten or eliminate this time? How much would this cost?
At first glance it seems there are a number of failures that you can’t do anything about; a power cut for example, or an internet outage. But generators can provide power for essential services and mobile connectivity can replace broadband in emergencies.
Also, it’s not always necessary to provide full redundancy for a device or service. Rather than having two of everything, perhaps you can provide a limited response to your main server failing using a PC pre-configured for the task and held in reserve. Does you support company hold hot spares in stock or would they need to order them in? The cost of holding a hot spare is often dramatically cheaper than the cost to your business of being unable to work.
Once you have the diagram, its annotations and the input from your technology support, you’re in a position to make some decisions based on risk and cost. Even if you then decide to do nothing more at this stage, you’ll have informed both yourself and your technology support of the priorities for your business and the equipment that supports it.
Is this something that you should do? Absolutely and it need not be more than 3 or 4 hours work, combining technical help with the expertise and knowledge of your business. If the worst does happen then it could be the best afternoons work this year.