Digital transformation may well be the IT story of the year, if not the decade. But while many enterprises are focused on how to make that transition, not enough are factoring in fundamental changes required to store and archive growing volumes of data on which they’re building the new digital enterprise. At least that’s the view of Matt Starr, CTO of storage solutions company, Spectra Logic. He says many enterprises on the transformation path are overlooking the need and value of digital preservation.
As McKinsey consultants observed in 2015, going digital means different things to different companies. But in general, digital transformation refers to the application of digital technologies to redefine all aspects of business operations and avoid becoming the victim of digital disruption. IDC predicts digital transformation will be at the center of corporate strategies for two-thirds of Global 2000 CEOs by the end of 2017.
When digital assets are the business
“Fifteen years ago, there were very few companies whose product was actually a digital product, but today for many companies the digital asset is what they monetize,” Starr points out. “How you now protect that, archive that, and store that is starting to become a challenge.”
When data is the business, it’s essential to be able to access the correct data quickly. “To do that you need to have a lot of metadata that describes what that dataset is, be it a SAP database that describes some point in time, or be it a seismic study from a plot of land somewhere in Dakota,” Starr explains. “That metadata collection isn’t always done well and it isn’t always engrained in the process.”
Degrees of magnitude
As enterprise storage blossoms beyond terabytes to encompass tens or hundreds of petabytes, the challenges grow by degrees of magnitude. For some businesses it may be essential to catalog different versions of data stored over time, such as in genomic research; for others it may be managing huge volumes of video content such as in automating a 24x7 cable network.
What constitutes valuable metadata may differ from company to company and industry to industry. It could encompass device identification, device firmware, rights management, and virtually any descriptive, structural or administrative information that helps in the retrieval, preservation, and relationships of data.
Is that an archive, or a backup?
As more and more data is collected, the distinction between backup and archiving it will become more crucial. “You get to a point where it may not be possible to backup a petabyte of data in a night; in fact you can’t,” Starr says.
“Administrators really should think about archiving data first and then about what they want to back up,” he advises. “If they archive data, they make two copies in an archive, one is a primary copy, and the second is a backup.”
That, Starr adds, “reduces the actual active data set that you need to backup; if you’re not touching 80 percent of data of that one petabyte, let’s say, why back it up once a week? Get it off to an archive, make two copies of it, and only back up the active 20 percent of your data.” And if you’ve accumulated the right metadata, data retrieval and recovery of the digital business will be much more manageable.
