How Copy Data Virtualization Helps Boost IT Efficiency

by Chris Preimesberger

Enterprises have two kinds of data: production data and copy data. Production data is the data flowing through the applications that power your business right now. Copy data is everything else. This is data produced by those same applications at some earlier point in time and later copied and retained by some other business system for some other business purpose.

Your backup system makes a copy of production, at some frequency dictated by your recovery point objective and recovery time objective requirements. If you have a disaster recovery system, it’s doing the same thing—day in and day out—as is your business continuity system. Developing applications? Your database administrators and application engineers will need still more copies, and nearly all companies create copies to support analytics, compliance and other enterprise use cases. Look hard and you’ll find systems making copies of your copies, putting you on a geometric data growth curve that makes the storage vendors happy while drawing valuable resources away from more strategic IT initiatives.

According to IDC’s recent Insight Report, untamed copy data will cost businesses more than $46 billion—that’s right, with a “b”—this year. Copy data typically represents 60 percent of enterprise disk storage systems, along with 85 percent of storage hardware spending and 65 percent of storage software spend, according to IDC’s recent white paper, “The Copy Data Management Market.”

The copy data problem was created by a series of data management decisions that made sense at the time. Organizations needed copies for backup, disaster recover, business continuity, dev/test, analytics, compliance and many other use cases. All were created for a good reason, but each without understanding what copies already existed. As enterprise data reached petabyte scale, the costs became so great that systems collapsed on themselves. The problem has become so acute that Gartner recently added copy data management to its 2014 Hype Cycle for business continuity and IT disaster recovery management.

Copy data virtualization starts over with a blank sheet of paper to take a holistic approach to the problem of enterprise copy data management. It solved the twin problems of proliferation and utilization, just as others had solved them in compute and networking: through the development of virtualization technology designed to manage the full lifecycle of data.

Where most data management technologies are designed to work up from the infrastructure, true copy data virtualization was designed to work down from the applications. This break with the past—the development of a data-management model that was application-centric and infrastructure-agnostic, rather than the other way around—is the key to understanding copy data virtualization.

The best copy data virtualization solutions capture data directly from production applications, including VMWare, Oracle, SQL, Windows, Linux, Microsoft Exchange, SAP and others. It will do so at the block level and in the native format of those applications, so it can be accessed immediately when needed. The system is smart enough to recognize the changed blocks since its last operation and simple enough for you to create application-specific service-level agreements in just a few clicks.

Having captured this application data, the system creates a single “golden master” of the data on whatever infrastructure you like—in the corporate data center, a remote location, or in a public, private, or hybrid cloud. After inception, this master can be updated incrementally forever, drastically reducing both the storage footprint of the data and the bandwidth required to manage its movement across the network.

Using this golden master of native-format application data, copy data virtualization solutions act as a storage controller, making an unlimited number of virtual copies available for use in support of any use case in any location. The system retains the metadata to instantly present a virtual copy of any application from any point in time.

Customers say they’ve been able to realize dramatic improvements in service-level agreements (SLAs) while radically simplifying their infrastructure and internal operations. It is not uncommon for the company to deliver these benefits while cutting the total cost of ownership for a customers’ data management infrastructure in half. Use case example: By replacing point backup tools and tape with Actifio’s copy data storage solution, Texas Wesleyan University in Ft. Worth, Texas, reduced recovery time SLAs from three to five days to five hours—an 18-fold faster recovery. The return on investment was impressive; the school saved an estimated $2.5 million in remote disaster recovery site costs and eliminated five-figure annual tape operations and connectivity costs.

Perhaps best of all, freeing essential application data from commodity infrastructure enables customers to move more aggressively toward the promise of cloud-based infrastructure, where the data from any application at any point in time is available instantly wherever it is required by the business.