Top 10 Tips for a Successful Data Center Design Project
- Published on January 6, 2011
- Written by MICKEY S. ZANDI, Ph.D.
Today’s highly competitive global and interdependent companies cannot tolerate their corporate data center to run with operations, availability, or resiliency deficiencies. With energy costs rising, organizations are looking for ways to reduce their total expense of operations. This cost-savings push demands right-sizing and consolidation of the enterprise assets and infrastructure, including the data center. The paradigm of doing more with fewer resources and smaller budgets is challenging to CIOs and compels them to look at various alternatives to address such demanding hunger for more computing power and storage. In parallel, the advent of modern data center transformation, through innovations such as cloud computing and transparent infrastructure, allow the organization to focus more on their core competencies.
Organizations build data centers for a variety of reasons. The latest key drivers include rising utility costs and limited space and cooling, all of which have created an environment of “must-take action” and caused many new initiatives to be approved by executive management.
Although data center managers have typically had to rely on “tactical fixes” to buy some time, the inevitable options in the take-action stage are to retrofit, design, and build a new facility or explore third-party options such as colocation, managed services, or cloud computing to address the current needs and augment the mission-critical infrastructure.
Laying The Design Foundation
Once your organization decides to build or retrofit an existing data center, you need to start the process of collecting, synthesizing, and exploring your needs and design specifications. Most importantly, executing such a complex project requires technology planning expertise, process standardization, and management, since the end-state design needs to be flexible, scalable, and efficient.
Just as with any design and build, data center design projects can be filled with mishaps: budget over-run, changing business and IT needs, technology changes, and many other unpredictable circumstances or hurdles. These challenges impact and slow the design portion of the project and put the project at risk – for potential loss of time and even revenue if the existing data center is out of space, power, or cooling. The lifeblood of the business, not to mention corporate reputation, could be at risk if the existing data center experiences downtime into “business hours” or if it becomes known by key customers or partners.
Many organizations have limited and over-committed resources. Planning and executing a complex design project adds a considerable risk to an ongoing shortage in staff utilization, efficiency, and production. Improperly planned and poorly designed data centers can have a profound impact on the business and result in significant lost revenue. The key to any successful project will be to understand your business and IT needs, plan, and engage proper competencies, methodologies, and expertise to execute such a complex task.
Design The Data Center With ‘Flexibility And Scalability’ In Mind
Companies that consider flexibility and scalability as the building blocks of their new design are starting the project off on the right foot. There are many parallel best practices between establishing effective design concepts and keeping the plan efficient and executable. The following top 10 tips are recommended to any organization looking to execute a data center design and build project:
- Understand your “application, storage, and network” requirements. Perform a complete discovery and needs analysis that maps business needs to IT resources.
- Understand your data center tier requirements. The first step for any organization in designing a data center is to determine the level of redundancy, resiliency, and criticality it needs.
- Develop your “phased” master plan. The design should be deployed in phases, and IT should work very closely with facilities to ensure proper sizing of the critical infrastructure components.
- Prepare blocking diagrams for all current functional areas. Apply these calculation plans to the new design to achieve the best location for the data center and support area.
- Develop a facility plan for the new data center identifying the location of equipment within the facility. This plan includes processors, disk, tape, network gear, power distribution, precision cooling, fire detection/suppression systems, raised floor system, under floor preparation, and cable distribution trays.
- Prepare cabinet standards and floor plan placement for IT equipment and distribution of cabinets/racks. Prepare cabling requirements for server distribution and develop workstation faceplate and cabling standards. Taking the time at the outset is critical to helping avoid and overcome the unpredictable mishaps that are likely to crop up later in the process.
- Design and lay out the main distribution frame. Prepare horizontal cabling testing and documentation standards. Research and validate your plans with a proven data center design best practices standards (EIA 942) as your guide to ensure success.
- Consider partnering with a technology services organization to minimize business impacts. Data center managers and IT staff do not need to go it alone. Look into partners to help support and streamline the process, but make sure you choose a firm that is experienced in design planning and executing complex engagements. Look for a company with proven methods and approaches that are focused specifically on data center design projects. Any partner should specialize in complex project management and be able to oversee all aspects of a design from assessment to specification and construction coordination and management.
- Provide site monitoring reference points for early warning and facilities trend analysis. Document, document, document. Track your facility, hardware, software, and network resources. Understand and document your systems interfaces, and don’t forget any applications and services.
- Begin with an end-state design in mind. Many athletes use visualization techniques to prepare for major events. For instance, golfers think about where they want the ball to land before they swing. In data center design, knowing the tier class, redundancy needs, and resiliency for all of your devices and technology is critical to making the design phase a success. Committing a mental picture to paper is a recommended step as well.
An organization should not expect that a data center design project will be easy and seamless, but following these ten best practices will certainly set the design team on the right path. Complex projects like this require carefully thought-out plans and a team of dedicated professionals willing to do whatever it takes to design a flexible and scalable data center for the 21st century and beyond.
Dr. Mickey S. Zandi, managing principal at SunGard Availability Services, is a recognized international expert in data center design and architecture, helping clients address key issues such as how to reduce total cost of ownership (TCO) in their data centers, how flexible infrastructure impacts data center operations, and why companies should consider new technologies to support their green initiatives worldwide. Zandi has more than 14 years experience in data center design and architecture, business continuity, and IT consulting services delivery, managing complex IT projects worldwide in industries such as communications, distribution and retail, education, banking and financial services, government, healthcare, IT services, manufacturing, pharmaceutical, and utilities. He is experienced in all phases from discovery of client needs to solution design and delivery through ongoing operation and methodologies. At SunGard, Zandi is the national practice partner responsible for data center consulting consolidation and relocation services.