• United States
by Dean Evans and Ryan Martin, Align, special to Network World

5 tips on how to optimize your infrastructure

Jun 03, 20117 mins
Data CenterSaaS

This vendor-written tech primer has been edited by Network World to eliminate product promotion, but readers should note it will likely favor the submitter’s approach.

While many enterprises have been able to creatively manage IT demand through the recession without much infrastructure change, exponential data growth is driving the need for infrastructure consolidation and optimization.

Here are five tactics that have proven to help IT transform the data center, taking into account current economic and financial conditions and the importance of delivering ROI while minimizing capital expenditures:

Tip No. 1: Leverage lower-tier data centers

Leveraging lower-tier data centers for resiliency is one of the most effective ways to lower costs because the higher the tier, the bigger the pocketbook burden. Many organizations with high-availability applications maintain redundant high-tier data centers. While these applications still require high availability, you can use lower-tier and lower-cost sites for non-critical applications and archival storage. In fact, this should be the case for any organization outside of certain government operations or those in the financial services industry.

Q&A: Uptime exec: Data center ratings aid cloud choice

When evaluating the option of moving to a lower-tier data center, your business needs and requirements for recovery times should be top of mind. Recovery times are independent of a data center’s availability or uptime, and provide flexibility for use of lower-tier data centers.

Cooling and operating a $2,500 server in a Tier-2 data center costs about $1,320 in electricity and facility operations compared to $1,870 for a Tier-3 center. That is 42% premium. As the electricity portion of the costs rises, the disparity between the two tiers will rise as well due to the differences in equipment power and cooling requirements.

Tip No. 2: Consolidate data center operations

Another path to optimization is to consolidate internal data center operations. The opportunity to significantly reduce costs and gain “green” credits by reducing your footprint is hard to ignore. Savings from such a transformation can range from 20%-60%.

However, the investment — both time and expense related — is just as considerable as the savings. For example, the Australian government anticipates spending $1 billon to achieve $1 billion in savings over the course of a 15-year project. That’s why it’s imperative to thoroughly lay out the approach and take timing and scale into account. Another example of this is Intel. With 100,000 servers in more than 97 global data centers, Intel started investigating its inventory as early as 2001, then began executing its plan in 2006. The plan is expected to take seven years to complete.

Tip No. 3: Choose low-cost relocations

Historically, major data center facilities are located within 100 miles of company headquarters. Accessibility near a major airport, technical talent and access to high-speed communications were once some of the main drivers determining data center locations. However, today’s drivers are led by access to large and inexpensive amounts of power, with small cities rapidly advancing as a top pick. Google is in the process of completing a number of new data centers in places like Council Bluffs, Iowa, The Dulles, Ore., and Pryor, Okla.

EXAMPLES: Apple, Google, Facebook turn N.C. into data center hub

For organizations hesitant to scout for new sites and negotiate direct benefits via tax and property breaks, special real estate developers are stepping up with prefabricated lease and build-to-suit options. One such development firm is Digital Realty Trust (DRT) whose recent expansion in downtown St. Louis was spotlighted in The New York Times.

Tip No. 4: Reduce the data center footprint

With the ever-increasing growth in information, there’s a constant need for more space. And yet access to capital will continue to be difficult for some time so you must understand and plan for alternatives to building or leasing an incremental data center.

Alternatives include everything from shared facilities to co-location and even cloud. Some not only displace capital expenditure spending, but also provide cost savings, flexibility and scalability not attainable with a new data center.

One interesting example is the holding company that was established after investment firm Lehman Brothers was dissolved in 2008. Faced with a major restructuring, Lehman Brothers Holding Inc. (LBHI) turned to an outsourced IT services firm to help determine the technologies that could create an infrastructure which would be quick to deploy, flexible, highly secure and cost-effective. Critical to LBHI’s success was its decision to “rent” infrastructure on a monthly basis rather than buy equipment. This allowed LBHI to save on capital investments while utilizing the services firm’s project-management, integration and managed-services expertise.

As the transition got underway, an on-site, private VMware cloud-computing platform delivered speed and flexibility for time-sensitive applications and data, while less critical applications and data were migrated to an off-site location. This hybrid approach maximized the benefits of both solutions — low-latency computing through on-site technology combined with cost savings from off-site storage. In order to ensure business resiliency, all locations were also replicated to a disaster recovery site.

Tip No. 5: Put SaaS and PaaS to work in the data center

Until recently, cloud solutions like software as a service (SaaS) have been characterized as having form over function, designed for a large population of users, and ideally suited for small to midsize businesses. However, as IT is increasingly tasked with tightly managing budgets and showing rapid ROI, SaaS must be moved closer to the top of the list. This model not only reduces infrastructure and support requirements, but helps speed time to implementation.


Platform as a service (PaaS) provides a couple of options. From an infrastructure perspective, development environments that use any of the platforms can be eliminated, reducing server hardware from the budget. From the application perspective, development teams can be grouped within PaaS enterprise licensing arrangements and have access to new environments in real time, reducing internal infrastructure middlemen and time lags. Finally, PaaS frameworks are open-source code and can typically easily be integrated with their SaaS counterparts.

These new models require the transformation of IT functions into a comprehensive service-based approach, and demand tight integration among application, facilities and technology teams to plan, design, migrate and operate data centers that accommodate a more flexible service delivery architecture.

Begin with a blueprint

Finally, realize that optimization is not a “once and done” opportunity. This is particularly true now because technologies and procurement strategies are changing at such a rapid pace. SaaS and PaaS solutions are just the tip of that iceberg as we move into the age of cloud-based everything and virtualized work environments.

In this new landscape, you must define an explicit roadmap and build in the flexibility to take advantage of new opportunities — such as implementing the latest energy management techs.

As they say, the best-laid plans can always go awry, but those plans are in vain without transparency into the organization’s current assets, from facilities to hardware and applications. Developing an accurate blueprint requires knowledge of the use of technology, the business and analysis of their interaction.

Read our other tips, from wireless to cloud computing

The rubber meets the road when the map is put into action. To ease the organization into the transition, and secure the buy-in for the rest of the plan, run a pilot to not only test the plan, but gain positive results. In this area, many organizations chose to turn to an external adviser to dig in and help undertake the transformation as well as build the case for a broader project that ensures the optimization process becomes a reality.

Dean Evans is managing director of IT strategy and management and Ryan Martin is director of IT strategy and optimization at Align. Align is a global information technology solutions firm, helping businesses move, improve, integrate and manage their entire technology infrastructures. For 20 years, the world’s leading firms have relied on Align to empower their IT — enabling business and IT to achieve alignment, and delivering solutions for business change and growth.