Spring Cleaning: Maximizing Your Space
“March bustles in on windy feet
And sweeps my doorstep and my street.
She washes and cleans with pounding rains,
Scrubbing the earth of winter stains.
She shakes the grime from carpet green
Till naught but fresh new blades are seen.
Then, house in order, all neat as a pin,
She ushers gentle springtime in.”
Spring has traditionally been viewed as a time for cleaning, dusting, sweeping, scrubbing, polishing, straightening up and organizing one’s living space. The timing of this ritual is not accidental, either: Typically, the onset of warm weather after months of bitter cold invigorates people to the point where they can take on tedious tasks (like cleaning) with gusto.
Similarly, it might be a fitting time for information storage professionals to tackle the arrangement of their own or their organization’s data, and come up with ways to better systematize their virtual area. After all, despite the leaps and bounds in storage technology in recent years, the amount of information has grown at a comparable level and, hence, space is still very finite. Here are just a few suggestions for storage professionals who want to clean house this spring:
Get Rid of Duplication
Scratch the surface of any company’s aggregate compiled data, and you’ll probably find duplication. Lots of it, in fact. This can be due to a number of factors: inefficient processes, convoluted organizational structure, end users who don’t know any better or (most likely) a combination of these working in concert to mess up your storage strategy.
While you may not have any control over the sources of these problems, you certainly have the power to react to them by eliminating superfluous data. The first step is making sure that the duplication isn’t necessary or desired on the part of the heads of the organization—perhaps they want their employees to save their reports on both their desktops and the servers. Of course, you can always make recommendations to them about this being a waste of space, but be sure to tailor those arguments to the intelligence level of a zoo baboon.
Once you’ve determined that the excessive data has to go somewhere, anywhere else, then you can start thinking about the next step.
Backup or Eliminate Data
I probably don’t need to go over the finer points of the latter with you: Suffice to say, make sure you’ve cleared eradication of particular pieces of information through all necessary channels. You don’t want some manager stopping by your cubicle after the fact to grumble that she really, really needed that file that you zapped into cyber-dust.
As for the former, determining that data needs to be backed up is only the beginning of the process. This brings us to the next and final recommendation.
Classify Data by Importance
There are many modes of data classification that pertain to backups. Naturally, the most mission-critical data—the kind that requires the “five nines” of availability—will be continuously captured on primary-disk systems. Loss of this category of information for a period of only a few minutes can cause profits to fall and customers to walk, and may even threaten the very survival of the business. Thus, this data is usually transmitted to multiple remote standby sites, which can recover it nearly instantaneously.
After this comes “vital data,” which—as the name indicates—is important to be sure, but doesn’t require the same immediate recovery time or accessibility levels (only three nines of availability, according to Horison Information Strategies’ data classification arrangement). Incidentally, most mission-critical information becomes vital data over time, sometimes in a mere matter of hours. Most of this stuff is kept on systems like SATA disks and virtual tape storage libraries, and replicated via snapshot copies and journaling.
The next tier is sensitive data, which refers to information used in regular, run-of-the-mill business processes. As with vital data, this kind is stored primarily on SATA disks, but massive arrays of idle disks (MAID) arrangements have become increasingly popular for this category.
The bottom layer of storage is non-critical data. This class of information is usually by far the largest segment of information stored, and often represents the greatest opportunity for freeing up space. It’s not highly sensitive or critical, so it doesn’t require highly developed security systems or immediate availability. Additionally, you’ll generally find the most duplication here. Where possible, this data should be copied onto a tape and moved to a remote physical location—out of site, out of mind. Few in the organization will ever need it, and if they do, it probably won’t be for any pressing matter.
Happy cleaning, storage pros!
–Brian Summerfield, firstname.lastname@example.org