3 Things That Matter in a Data Center

Updated: February 08, 2011

What - Differentiating the Technology

While you might think "a data center is a data center" or at least that all Tier 2 or 3 data centers are alike, that's only partially true. Sure, many providers tout familiar features such as N+1 UPS, redundant networks, or "five nines" availability. Then there are the terms that most marketing people have abused when describing data centers such as "scalable" and "robust". Decision makers are having an increasingly hard time separating the various offerings and assessing providers. When all of the data centers are claiming the same thing, using the same language, how can you reasonably tell the difference? More importantly, how do you find the right fit for your unique business requirements?

There is no doubt that what's "under the hood" at a data center is critical. The technology infrastructure as well as the physical infrastructure should be carefully reviewed to ensure they meet with your organizational goals and objectives. The differentiating factors, however, might be difficult to uncover. Comparing server models, processor speeds, or virtualization technology will probably leave you with little differentiation between providers. Those differences are also too easily erased by a simple hardware upgrade. What matters is how the data center provider implements and wields the technology.

When your data center provider focuses on purpose-built performance, maintainability and security, you'll begin to find the differences. Do they own their building or space? This might not seem important, but tell that to your IT team when they're trying to push through a 3rd-party building landlord at 2 am on the eve of your product launch because your colocated servers are down. Within the data center itself, do you see innovation in design such as individually cooled cabinets that completely bypass issues associated with hot/cold aisle management and data center "hot spots"? Does the provider pay attention to details that may only become important sometime down the road - like color coding their wiring schematics to match the sheathing colors on the actual cables? At GlobeOp Managed Services, we've done all these things and more. We've done them because we rely on our data center as much as our clients do.

Keep in mind that the technology, while more easily bulletized and compared, is NOT the only thing to consider when selecting a data center or colocation provider. In fact, technology may not even be the most important thing you consider given that the overwhelming majority of issues in a data center are caused either by human error or process problems.

Who - People Behind the Scenes Make the Difference

When you get right down to it, "who" operates a data center is every bit as important as "what" technology they are using and "how" they manage it. Consider how businesses struggle with their own IT operations. Between 2007 and 2009, 48% of corporate data breaches were attributable to company insiders. Sixty-one percent were discovered by a 3rd party and 96% were deemed avoidable through simple and intermediate controls. Your selection of the optimal data center with expert professional management and tactical resources can address these kinds of issues and more.

You should be assured that the data center personnel, from executive level, through management, to the operations staff, are highly trained in the disciplines required to build, manage and evolve the data center while maintaining state-of-the-art security measures. This means the executive leadership should believe in the underlying principals of accountability, continuous improvement, and strategic investment. Management must be both knowledgeable and empowered to make the best decisions for clients and be nimble to manage risks before they manifest to issues. Operational staff ought to be appropriately allocated - with clear separation of duties - and comprise a base of expertise that powers improvement, technological innovation, as well as rigorous controls over risk areas and security.

At GlobeOp Managed Services, our pedigree in the financial services industry has mandated high performance teams in all areas. Accountability starts at the top with us and runs transparently to the foundation of the business. Our technical staff maintains proficiency to an in house curriculum of technical subjects to stay up to date with the latest innovations while possessing over 110 separate third party professional IT certifications from Microsoft, Sun, Cisco, IBM, Citrix and more.

We also maintain an unparalleled depth of expertise in data management planning, execution and security. These data management practices were born of a need to support our own financial middle and back office products for over $100 billion in alternative investment assets and to support business continuity and disaster recovery scenarios for the same.

So when you look at your business IT needs, consider the team that will tasked with meeting them and ask the tough questions. With top-to-bottom support, does the
data center have personnel highly trained in networking, security, data management, hardware, infrastructure design, and the hosted or managed applications that keep your business running?

How - The Processes That Fuel Performance

As IT infrastructure, business applications, and compliance issues grow more complex, the possibility of an error, technical issue or breach grows also. How your data and
infrastructure are managed is a direct result of the people at the data center as well as the processes they use to ensure reliability, security and continuity. The processes should be proactive and built around managing and mitigating risks - not just responding to adverse circumstances after they happen.

It's clear that process is a key element of a well-run data center. How do top data centers optimize their processes?

  1. Continual review and improvement of processes by qualified staff
  2. Organizational focus on process output as a top priority
  3. Deep experience to draw from and numerous real-world situations
  4. Investment in internal and independent auditing

A challenge for most data centers is that their processes are not developed enough to produce reliable output or fluid enough to adapt the changing data center environment, applications, and security threats. Again, because most data centers are built as rental profit centers first, and performance centers second or not at all, they start with a set of standard processes and practices, but they do not have the staff or expertise to continually evaluate and improve upon those generic measures.

Process maturity must be a core value of a data center if it is to offer the most sophisticated services. At GlobeOp, because the data center infrastructure and team must support the enterprise business requirements for trading over $100 billion in financial assets, the continual refinement and evolution of processes is mandatory for the data center and a priority for every single professional within the organization. EVERYONE is accountable for compliance and improvement, not just one compliance officer or a single department. Specifically, GlobeOp has 60 directors and 400 managers, all focused on top-to-bottom quality and security as prime directives.

Our commitment to process - and more importantly process output - is demonstrated in several ways. For example, over the past four years of SAS 70 Type II audits, GlobeOp Managed Services has not had a single corrective action. In fact, auditors have not reported a single negative finding.

Best practices at GlobeOp go beyond audit achievements. Running our financial services sister businesses out the same data centers we offer to you means our own risks are bigger, and thus so are our mitigations. That's why you'll find we go above and beyond the norms doing things like:

  • Twice per year professional cleaning under our data center raised flooring to prevent particulate migration
  • Per-battery automated terminal voltage monitoring of our UPS strings with daily manual verification by a trained technician
  • Rigorous separation of duties for technical staff to ensure expertise is deep in each area of our operations, to prevent our team members from becoming over-allocated, and to maintain a focus on both the forest AND the trees.
  • Convening our "post-mortem" review board for every process development and incident resolution occurrence in the data center. We include review members from sectors not involved in the issues to garner objective validation in every case.

Related Categories
Featured Research