Published: 01 Oct 2009
Cloud storage is coming and it will undoubtedly bring benefits, but there may be residual effects to consider.
Before anyone accuses me of being a Luddite or some sort of cave dweller, let me say that I'm a great fan of innovation and I love technology. But at the same time, I tend not to jump on a new tech bandwagon until it takes a few spins around the block. To put that in storage terms, while I believe cloud storage will play a part in the future of a lot of storage shops, I also think we should put the brakes on that bandwagon a bit and consider all of the ramifications of the new storage world order it promises.
The main concern I hear from storage managers is that they're reluctant to ship their data off to a cloud provider's site. That's worrisome, for sure, but I think that hurdle can actually be overcome relatively soon. Plenty of shops have been shipping data offsite for years by using third-party disaster recovery services. But using a third-party facility to store a third or fourth copy of your data for emergency use is a lot different than using it for primary storage or even backup.
There are a number of other real issues that have gotten some air time but probably haven't been examined as closely as they should be, like the need for faster and wider communications pipes. Is it possible that the savings gained from using cloud storage could be offset or even outweighed by the need to build a better telecom infrastructure?
More perilous, I think, is the commoditization of these storage-related services. A cloud-based storage service -- no matter how much customization it might provide -- isn't likely to be as tailored, tweaked and tuned to a company's specific needs as an in-house setup. But the relative low cost of these services will be appealing to management, and the urge to bypass the expertise of the resident IT crew will be tough to resist. The traditional role of IT could change dramatically, turning into what could amount to a clearinghouse for external storage services. Again, staffing and capex savings might be enticing, but what happens when that dwindling expertise is needed again? It's likely that those special skills and expertise will have also drifted off into the cloud. It can be argued that similar evolutions have happened before and IT has survived intact. In a lot of companies, application development teams have evolved into off-the-shelf package evaluators/customizers, with less and less actual development and coding being done in-house. But storage requires different care and feeding. Given the wide variety of storage solutions available and the way they can be applied to unique environments, I don't think any kind of mass commoditization would be beneficial or realistic.
Which brings up another point: If a company's storage moves offsite and into the cloud, it could affect more than just the profile and configuration of its data centers. There could be a big shift in the design and manufacture of all storage systems. Cloud service providers don't use the same storage systems that corporations do -- that was tried back when storage as a service was first unveiled approximately 10 years ago. The economics of carving up a big enterprise-class array for a number of clients just never worked out. Today's cloud vendors tend to cobble together their storage systems into more economical and utility-like configurations. Those setups, which would seem rather unsophisticated compared to enterprise storage systems, might work well for delivering storage from the cloud, but they're not likely to suffice as in-house systems where performance may be paramount.
So as cloud storage grows, will storage vendors shift their efforts to building systems for cloud providers? In a recent op-ed piece in The New York Times ("Lost in the Cloud," July 19, 2009), Jonathan Zittrain, a Harvard Law School professor, expressed some apprehensions about cloud computing and how it could affect innovation: "But the most difficult challenge -- both to grasp and to solve -- of the cloud is its effect on our freedom to innovate." Although most of his discussion focuses on PC users, Zittrain is most concerned with software development falling from the hands of the many into the hands of a few service providers. With software the main differentiator among storage systems, a similar scenario could unfold relating to the development of new storage products.
Similar thoughts were expressed by Wikibon.org's David Floyer in a recent Wikibon newsletter, but more from the perspective of storage developers. "Wikibon predicts that [in the next decade], total spending on external IT services will exceed the declining traditional IT budget in most shops," Floyer wrote. He went on to describe how the attention of storage and other IT providers could (or should) shift, and added, "The traditional model of selling to IT organizations will be a declining market."
I don't know if Zittrain and Floyer raise issues about cloud storage and providers that will never come to pass, or if they're just a step or two ahead of the rest of us (or me, at least). In the end, it's the storage purchasing patterns of corporations that will go a long way in determining if the pictures they paint will develop as described.
BIO: Rich Castagna (firstname.lastname@example.org) is editorial director of the Storage Media Group.
- Cloud Storage for Primary or Nearline Data –SearchStorage.com
- How to get the best value from Office 365 –ComputerWeekly.com
- 3 Common Cloud Challenges Eradicated with Hybrid Cloud –SearchStorage.com
- IT Project ANZ: Managing datacentre growth –ComputerWeekly.com