The Building Blocks of Data Center Transformation

The heart of a company's technology infrastructure, the data center stands as the single most important asset to a bank's ability to conduct business. Data centers are the foundation upon which so many transactional and service capabilities are built.

The data center is the hub for the increasingly IT-driven business that is banking. And while certain aspects of data center management -- energy efficiency, physical space, electronic capacity and security -- will always be paramount, the data center constantly is in a state of reinvention. And it's not just about raised floors and cooling requirements.

But change in the data center often is incremental. While new technologies may influence an executive's desire to follow the bleeding edge in data center management, enterprise IT is charged with keeping everything running.

"There's all sorts of innovation going on" in data center technology, says Joseph Tobolski, senior director, cloud programs, for Accenture Technology Labs, who is based in Chicago. "A lot of enterprise data center operators, however, are not incented to make radical change because they're charged to keep the machines running and have no errors."

And change isn't easy. "It's difficult in a large, heterogeneous IT environment to make wholesale change," Tobolski says.

Banks, like any other IT enterprise, will always have to consider hardware and efficiency as part of the data center conversation. But running an efficient and effective data center requires more strategic planning than that, suggests David Rutchik, a partner with advisory firm Pace Harmon, which has offices in Washington, D.C., and San Francisco.

Building Blocks: Automation, Virtualization and Cloud

Boiled down to specific areas of interest within the sphere of financial enterprise IT, there are three hot topics surrounding the data center, according to Rutchik: automation, virtualization and, ultimately, cloud computing. "Those three things all create more flexibility but fundamentally create reduced costs," Rutchik says, adding that hardware also remains near the top of many conversations, particularly when it comes to power and cooling.

According to experts, the thing to remember when transforming data center operations by enabling automation, virtualization and cloud computing is that the end goal is about improving IT service and delivery as much as it is about cost cutting.

Automation eliminates menial IT tasks. As in the manufacturing industry, automation in the data center is a means to reduce staff time devoted to monotonous, often repeatable requirements, such as patching or configuration, that can be handled by a machine. Automation allows companies either to shift the human workload to other, more value-added areas or reduce the overall need for manpower.

Virtualization frequently provides the enterprise with a means to deploy a controlled desktop environment to employees from a central location, often using a solution such as Citrix (Fort Lauderdale, Fla.) or VMware (Palo Alto, Calif.) to emulate a Windows environment. This improves standardization and reduces IT support and maintenance costs. But virtualization also can be implemented at the data or storage level for the sake of provisioning, disaster recovery and enterprise efficiency. It allows companies to parse computing workloads across servers and more efficiently utilize computing resources.

The cloud takes this one step further. Cloud computing is all about sharing resources, improving scalability and agility, and reducing costs while improving uptime. Put the three together and the data center starts to work for itself.

"This is not to say that the IT department goes away," Accenture Technology Labs' Tobolski stresses. "But it needs to be focusing on services, not trouble ticketing and workload processing, which can be done by a machine."

Getting the Cloud Off the Ground

These days it's difficult to talk data centers without getting into a conversation about the cloud. Popular among consumers and more than a buzzword for major technology players such as Amazon, Google and Microsoft, the cloud sits as sort of an ideal for shared, on-demand resources. But within the highly regulated banking industry, in which data security is a mandate, the public cloud is not the ultimate solution for all of an institution's data needs.

"There's also an assumption that the cloud is cheaper," Pace Harmon's Rutchik adds. "Everyone thinks it's going to cost less, but that's not necessarily true."

While the public cloud is not an option for most banks, Rutchik says, it is an example of what's possible. Within banks' data center operations, whether in-house or outsourced, there is discussion and adoption surrounding private cloud solutions. With a shift to the private cloud model still a way off for many banks, however, Rutchik sees virtualization as a first step toward a data center in the clouds.

New York-based Citigroup ($1.9 trillion in assets) is among the banks that have completed private cloud projects. In 2010 the bank worked with IBM (Armonk, N.Y.) to automate the provisioning of resources to developers in a cloud environment. "We saw this as an evolution of IT infrastructure that goes from a dedicated world where servers are built to order, to a virtual infrastructure in which the focus is on improving the utilization rates of our infrastructure," says Graham Hill, SVP at Citigroup.

An IT Transformation

Banks are among the biggest spenders on IT, Pace Harmon's Rutchik notes. But spending on the right technologies remains mission-critical.

Achieving a private cloud-like environment, however, requires a shift in perspective, Accenture's Tobolski adds, pointing out that private clouds don't come in an out-of-the-box, plug-and-play option. "It's important to reiterate or emphasize that if you want a private cloud, you have to build one -- you aren't going to buy it from anybody. The subtlety is that it's really an IT transformation story," Tobolski says.

"If you are a typical enterprise CIO, your users have access to technology and capability that used to be the purview of only big IT," says Rich Walsh, president, document archive and repository systems for New York-based electronic content management provider Viewpointe. "Going out to a cloud service provider and spinning up thousands of servers, or getting a bunch of software packages or really this whole notion of employee-powered IT, really disrupts the status quo."

If a bank wants to leverage virtualization and the cloud, it needs to change its approach, Accenture's Tobolski says. "You have to look at IT in a holistic fashion if you're going to achieve cloud-like capabilities," he contends. "And that is bigger than just the data center, bigger than just the kit. It's fairly extensive."

IBM fellow and chief systems engineer Guru Rao, however, says it's important to ensure that the right hardware is being used for the job. "Big and small companies should think holistically about workload," he says. "Pick the right systems, the right design, the right hardware, and tune it to the task."

But while the data center used to be all about nameplates and exciting big-vendor solutions, Accenture's Tobolski argues, the shift in focus to an evolving data center environment must come from within. And it must be focused on adding business value. "There's also probably a little bit of admiration and envy going on," Tobolski notes, suggesting that consumer cloud services such as Amazon Cloud Drive, Dropbox or Google Docs represent something of a holy grail for bank data centers.

The experts agree that the end goal of cloud, automation and virtualization projects is efficiency and flexibility, and shaping IT to reflect and support that agility. "If there's any trend, it's toward being more efficient," according to Viewpointe's Walsh. "I don't know if transforming the data center is the trend. But I think it's trying to be more efficient about your current environment."