August's Northeast blackout shone a spotlight on disaster preparedness, and in doing so gave us a way to gauge how far we've come in the post-Sept. 11 era. The early reviews: Progress, but not enough.
In July, I suggested that compliance might be the killer app that would bring storage networking solidly into the infrastructure, and storage managers firmly into upper management corps of IT.
I think I was on to something, but reflecting on the events of the last month, the broader question seems to be how to build data protection and retention into enterprise computing, not bolt it on. Sadly, that's not a new question, even if it has a fancy new name--information lifecycle management. But it definitely has a new urgency.
We now live in a world that's interconnected in more ways than any one of us can know. An angry peasant, an incompetent bureaucrat, an indifferent employee, an aging piece of equipment--any one of these and more can initiate a wave of disruption that will produce disastrous consequences in less time than it takes to read about it. Maybe these events can be prevented, but long before we live in a world free of such causes, we need to live in a world where we can at least recover quickly from their effects. A small, but significant, component of that capability lies in your hands.
Today, the ability to automatically transport data to multiple, secure and separated repositories and retrieve that data as needed is at an all-time high. The obstacles to doing so are increasingly nontechnical: failure of will, failure of vision, archaic organizations and even just plain incompetence. Storage managers, more than anyone, know what's doable and how much it will cost. So as you work with your CIO on next year's budget, stick your neck out of your collar a bit, clear your throat and ask how the business plans to deal with future disasters before they happen. Not only shouldn't you take "nothing" for an answer, but you should also reject vague pat-on-the-head answers that lead nowhere.
As for the bucks to back up plans: Attack on both fronts. The cost of doing nothing is clearly higher than doing something effective for those who are the victims of a disaster, and that seems to be a large club these days. What's less obvious to your colleagues is that business as usual can cost less, too, freeing up funds to do better. Frequent Storage writer and all-around smart guy, Marc Farley, will explain in-depth next month, for example, how the combination of Serial ATA, RAID and storage area networks (SANs) completely changes the economics of disk storage while improving reliability. That's a change you can make in future purchases with little additional effort.
Other modernizing changes, such as using IP SANs for scattered file servers and then replicating them asynchronously to disaster recovery sites, may take some time and people, but selective evolution of your infrastructure is certainly called for. Disaster preparedness can't be a wish in these times. What some people thought of as "ain't broke" obviously was in the clear light of a blackout. Now, can we fix it?