12_tribes - Fotolia
A new ritual has developed over the past few years where vendors -- and their PR firms -- wrap up the old year and embark on the new by pitching their lists of predictions about the future of data storage. Most of the prognostications look a lot like those vendor surveys that typically suggest that the world's data centers will go to hell in a handbasket if people don't buy and deploy their products ASAP. Their predictions read more like a sales teams’ wish list than any kind of insight into the future of IT.
And, of course, there are always some vendors that seem to be using second-hand crystal balls when they predict not just the obvious, but things that are already happening in the data storage arena. One vendor prophesied, "2015: the year of software-defined storage, flash, and virtualization." Wow, they went way out on a limb with that one, huh?
Another inched its way out onto that same branch with the gutsy call that "in the coming year, enterprises will continue to demand IT infrastructures that allow them to rapidly and efficiently deliver quality services." That was submitted by a PR flack on behalf of his client, which, I hope, had no knowledge of this bold look into the future of data storage in 2015. (Nonetheless, I'm sure many IT directors will be relieved to know they won't have to deliver crummy services slowly and inefficiently in 2015.)
Another vendor made a good observation about how flash storage vendors will have to find more ways to differentiate their products because plain old performance won't be enough to establish a product's value anymore in a very crowded, competitive market. But then they kind of lost me when they said "companies will see that driving an orange Ferrari while wearing a blindfold is cool for a short time, but not very bright in the end." I have a red 1994 Honda Civic, so I'll have to take their word about the orange Ferrari and the blindfold.
Some other predictions may sound pretty good at first, but when you think about what they really mean, they don't work. Like this one: "The days of buying storage array hardware and then being pressured into buying mandatory bolt-on software, with costly maintenance charges, are now disappearing." That could very well be true, but does that mean we'll buy storage software and then be pressured to pay a premium for the hardware to run it on? Guess that's one of those glass half empty, half full things.
Some vendors seem to have more cooperative Ouija boards or Tarot cards. Avere, for example, focused on the hybrid cloud with the safe yet prescient remark that "The cloud is becoming a must for almost all businesses, but they still need to balance on-prem security/policies with the convenience of the cloud." That's more practical than daring maybe -- and a rational prediction to say the least -- but it went on to call for "flexible cloud-based services" so "companies can on-ramp and off-ramp data migrations between providers." Think of that; being able to easily move data from cloud service to cloud service or use one to back up the other. We're getting there and Avere predicts that "It will be as easy as switching channels on a television with a remote control."
Those are spot-on, but they're not really predictions about what will happen -- more like what should happen. Despite all the blustering by vendors, analysts and other pundits about how the cloud is big, and how cloud is everywhere, and how everyone is doing cloud, there's the hard truth that it's just not happening yet on that scale. Notice I said "yet" because it's very likely to reach those proportions … but Avere's predictions will have to come true first.
Another set of data storage predictions that rang true to me came from Tarmin, which said there might be a silver lining to all the data companies have been hording, and they'll actually start to look at ways to make money off all that stored stuff. Tarmin goes on to predict that "this in turn will drive the need for data centric approaches that better enable organizations to build data management strategies based on the value of the data." And that's the part I like, the part that's really perceptive. Data just has to get smarter.
I've been stumping for smarter data for a while. We have the ability to make data a lot more intelligent on its own by tucking some useful metadata into it. Object storage is particularly adept at this, but that capability seems to be stuck between two tech worlds that -- for silly reasons, no doubt -- seem to be resisting the movement to make data brainier.
Smart data has to start at the application, because that's where the business intelligence lies. But not much appears to be happening in that realm. So, if the apps can wise up a bit, they could pack that metadata into the files. But then what? There's not an awful lot happening on the back end either, with storage management systems that can take metadata-based marching orders and act accordingly to delete, copy, restrict access and so on. But there are a few bright spots to suggest that some progress is being made. DataGravity, a startup array vendor, recently debuted its Discovery Series of "data-aware storage" that analyzes and classifies the data it stores so that it becomes more meaningful and allows more automation of data management.
Whether vendors' predictions are profound or preposterous, one thing is certain: With flash, cloud, hyper-converged, software-defined, scale-out, object, erasure coding and so on, it looks like you'll be pretty busy sorting out storage stuff this year.
About the author:
Rich Castagna is TechTarget's VP of Editorial/Storage Media Group.