Andres Rodriguez is the CEO and Founder of Nasuni, a unified storage company that serves the needs of distributed enterprises. Previously he was a CTO at Hitachi Data Systems and CTO of the New York Times.
When it comes to storage, IT professionals’ biggest challenge is figuring out how to serve an increasingly distributed enterprise. The problem is, as New York Times columnist Tom Friedman has memorably written, “The world has gone flat.” This trend is forcing IT to deploy infrastructure to a hodgepodge of remote locations with a broad set of requirements. Some locations may be as large as headquarters; some may be a half a dozen people with flaky network access.
Challenges of Distributed Locations
But everyone at these locations needs one thing: access to data. Being able to provide global access to data while delivering a uniform level of data protection is a formidable challenge in itself, and it’s compounded by the suddenness with which this need arises. In most organizations that are growing quickly, whether organically or through acquisitions, IT has a decentralized infrastructure. When the first disaster happens, however, and IT at HQ cannot get a handle on the situation or bring everyone back up quickly, the CIO is instructed to figure out how to make sure it never happens again.
Are Security and Backup At Odds?
Data protection really has two meanings that stand at odds to one another. The way IT protects data against loss is to make additional copies. But when it comes to protecting data against theft or an accidental leak, IT reduces the number of places where data can be accessed. There is an intrinsic tension between our need to secure and our need to make sure data is sufficiently backed up. Standards, like ISO 27002, capture this tension by requiring data security policies as well as business continuity policies to protect against data loss.
The way around this tension is to encrypt all data before it goes to backup and then make copies of the encrypted data. As long as the keys are kept under tight controls - and copied - it becomes possible to make as many copies as needed without compromising security. Modern backup software supports encryption to ensure that no copies of data ever leave the data center unencrypted. This can work well if your organization only has one or two locations that require IT infrastructure; however, the distributed enterprise creates an additional layer of complexity because data is being created in many locations and often it needs to be accessed from every location.
What About Cloud?
CIOs have started to look at cloud services as a way to deliver uniform infrastructure globally because the cloud enables IT to deploy distributed resources while retaining complete central control. Cloud providers offer an ever expanding menu of choices for IT, giving departments new ways to consume their global toolkit. Cloud projects range from transitioning complete application stacks such as email into Software as a Service offerings, to connecting layers of your existing infrastructure (Infrastructure as a Service), be it components of storage or the network, to the cloud directly.
Rather than having systems that are independently managed at each remote location, the cloud gives IT the ability to manage all enterprise data from HQ while simultaneously allowing resources to be deployed globally from its global backbone.
These benefits are not unlike those that arise from data center consolidation, but there’s a big difference: the best cloud providers’ massive infrastructure is far superior even to that of the largest enterprise data centers. No enterprise data center can approach the kind of availability, redundancy and replication across multiple geographies that a storage cloud like Amazon S3 can provide. At the same time, its usage-based cost model makes it far less expensive to deploy than data center consolidation. The cloud is data center consolidation without the data center.
Cloud is Consistent
Cloud infrastructure allows IT to deliver a uniform level of service anywhere in the world. Data is consistently encrypted and replicated to a common back-end that in turn is available to every other location. By allowing IT to tightly regulate access to what is in essence a global private distribution network, data can be both available and fully protected across the distributed enterprise.
However, it’s important to note that, when it comes to storage, the largest storage clouds are essentially raw components, much like the commodity disk drives that enterprise storage systems vendors like EMC and Netapp use in their storage arrays. On their own, cloud storage services like Amazon S3 and Microsoft Windows Azure Blob Storage cannot serve as enterprise storage solutions any more than could a large shipment of Seagate disk drives. Cloud storage needs to be integrated into a larger, smarter service by a new generation of systems vendors who can tap into the revolutionary potential of cloud storage to provide true Infrastructure as a Service.
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.