Skip navigation

Data Center Strategies: Road Map To the Perfect Strategy

Don’t let best practices, or stories of super-efficient designs cloud (pun intended) a data center strategy that should align with the business. In the finale of our Data Center Strategies guide, we examine the process of evaluating the buy vs. build decision.

A data center under construction in Silicon Valley.

With the context of knowing how data center technologies have advanced, who the stakeholders are inside the company, and what best practices have yielded, a data center strategy starts to come in to focus.

Action --> Evaluate capacity, cost and capabilities to build an optimized data center strategy. 

Own vs. Lease

The implication of leasing a data center instead of owning one comes down to the benefits of cost savings, flexibility and scalability and finding a partner worthy of trust. Uncertainty is the major deterrent to outsourcing, and providing for IT is nothing to be uncertain about.

A small number of businesses are aware that a wholesale lease is possible and available in many locations. Wholesale data center services present a unique opportunity for a variety of businesses needs. The wholesale provider focuses on real estate, efficient infrastructure and energy, which deliver flexibility and scalability for their tenants to create a long-term home for their infrastructure to grow.  Wholesale is an attractive option for enterprises that are looking to quickly expand their footprint of existing company-owned facilities without the risk of financing and building another facility.  It is a flexible choice that can deliver a greater value than colocation when looking to outsource.

No More Forklift Upgrades

Part of the pain of dealing with technology advances is the forklift upgrade – the need to upgrade or retrofit a facility that is out of space or can no longer deliver the power and cooling needed, a scenario that can occur in either company-owned or provider data centers . For organizations used to migrating from one solution to another to house their IT, the wholesale data center opportunity can offer a longer term solution, by applying vertical scalability.

The wholesale tenant is buying space that can be expanded as demand dictates, and if the provider data center has engineered it, power can scale as well. Put another way – you can optimize almost every aspect of your IT, why not optimize your data center instead of planning and budgeting for yet another move? While the best data center is one that doesn’t have to be built, many would amend that to be one that you don’t have to move out of every 5 to 10 years.

Two essential factors frame an effective data center strategy

1. Scalability

What scalability truly means to a business has evolved and grown along with the advances in technology and data centers.  The advent of virtualization impacted the context of scalability to shrink the server footprint, yet increase server density.  An increase in IT demand, whether through Internet businesses serving millions of end users or enterprises with increasing dependencies within IT, meant trying to scale energy consumption. Growing energy demands put a focus on efficiency, alternative energy resources and the financial responsibility for powering the data center. Vertical scalability implies that adequate power is installed in the data center for today’s forecast, and the ability to scale total facility power capacity exists. This makes a future upgrade possible without completely starting over with power equipment and utility feeds. Having a strong foundation like that built in to the design of the facility effectively doubles the life of the data center.

2. The Economics of Energy Efficiency

If building a data center or optimizing an existing one, energy prices and sources should be evaluated to determine the impact on capital expenditures up front and operational spending on the utility bill. For an outsourced solution, evaluate the provider's PUE and their ability to provide energy efficient power and cooling systems.  Achieving a lower PUE does not always equate to a higher cost of data center operations.  A net cost savings over time can be seen through implementation of high-efficiency infrastructure to power and cool a data center. Any level of power capacity can have efficiencies built in to how the data center components operate and save money.  A larger power draw has a greater opportunity for savings.

Financial Modeling

Most discussions investigating the true cost of owning versus leasing focus on the advantages of getting rid of capital expenditures (CapEx) in favor of operating expenses (OpEx). What is really being referenced in these discussions is cash flow timing differences, and not the accounting definitions of CapEx and OpEx.  When owning or building a data center, there is a large up-front capital investment made that is then depreciated over its life. Alternatively, if you receive data center services from a provider, you are paying a monthly amount and at the end of the term you can either leave the data center space or renew.

The true costs involved in up-front purchases versus regular payments to a provider are not mentioned in the typical cost analysis. Consider how the cost of capital, return on capital, tax savings and other economic incentives can contribute to the financial analysis. Evaluate direct, indirect and overhead costs to get a true picture of the total cost of ownership.

Don’t let best practices, or stories of super-efficient designs cloud (pun intended) a data center strategy that should align with the business. With a spotlight on resiliency, environmental and energy impact and most of all cost, the best strategy is an insightful, aligned plan to drive the business.

The entire Data Center Strategies white paper can be downloaded here.

DCK Guide to Data Center Strategies Home

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish