Disaster Recovery Plans (DRP) are predicated on systems and procedures working successfully together. If you're fortunate enough to have a plan, how do you keep it alive - because they seem to 'die' from a range of causes. In fact you could call them the 'fragile flower' of the computing environment. Which is kind of paradoxical, given that they are supposed to be your 'robust response to the most dire and unexpected events'. Actually that's the answer. Disaster Recovery Plans are (primarilly and often exclusively) designed to deal with disasters, and not designed (usually) to maintain themselves 'in-between' times.
It's not so much a disease that 'knocks over' DRP's but 'predation'. If the DRP includes hardware resources the sight of those idle 'ready to go' machines is an fatal temptation to the hard pressed organisation . But of course the 'fatal' consequences don't become apparent until you need the DRP to respond to some unpleasantness. But's that's not to discount the effect of 'drift' where your DRP isn't kept up to date with your changing business, or simply loss of knowledge about how to run it when the time comes. Sure you've documented the procedures, but where did you put that document - and has it been burnt/turned to pulp by floodwater/or sitting under several tons of highly compressed reinforced concrete? In fact, for that matter, where are you, and what particular condition are you in to co-ordinate things?
Is there an answer to the problem of keeping the DPR 'alive' and ready in your organisation? Perhaps, but the solution is not what you'd expect. If you turn over part of your DPR site to production activity (which is heresy in traditional DPR-land) you make it part of the 'active' landscape. In a Disaster you simply stop that aspect of production and switch the resources (and potentially the people) over to the 'Disaster Role'. I might add that of course we're talking about a physically remote site. I'd argue that it would be good to have the people using that part of the system there as well (and have them trained in running the switch-over), but it's not essential. So how do you stop that capacity from being turned from 'discretionary production' to 'essential production'. Well it's the same issue as the poaching of equipment in the first scenario - it has to be managed. But there are advantages in putting the DRP in the 'active landscape', and that's what I'll expand on shortly. And yes there will still need to be 'duplicate' embargoed equipment and systems within the DRP site, which means that we are talking about a complex solution.
Comments