Welcome to the Data Center blog. Recently I gave a webinar on Data Center Trends and one of the topics I covered was the concept of holistic design. I thought I'd make that the feature of my first blog post.
Whether upgrading a Data Center or building one from scratch, Data Center managers typically have two main considerations on their mind—cost and reliability. However, there is a lot more to the puzzle, and it starts with the acceptance of an often overlooked fact: A Data Center built only for the requirements of today can be a Data Center destined to fail in the future.
A Data Center cannot just serve existing needs—it has to meet the projected demands for the future. Unfortunately, upgrades have often been completed with a short-sighted "fix it" mentality where Data Center managers strive to just fill in gaps without a true 360-degree perspective. Cost has often been measured by how much it will take to implement a solution, rather than how much it will take to operate and its total effect on the facility. For example, Data Center managers often narrowly focus on the long-term benefit of a solution (and probably some hype), rather than asking themselves what the new solution will cost to run on a daily basis, if it truly offers a return on investment (ROI), whether ramp-up time is required for employees and how exactly a solution fits within the long-term goals of the business.
The good news is that Data Center managers can move from a "fix it" to a "build it" mentality through holistic design principles. With a holistic approach, the two philosophies that anyone upgrading or building a Data Center should live by are: Think about the outcomes and Bridge the gap among actual business requirements, IT and facilities. In other words, a holistic approach looks at the outcomes of every aspect of the design with input from all stakeholders. Some of the key holistic design considerations include:
- Multiple Tier Levels—Which systems needs Tier 3 availability or better, and what is the impact that downtime of these systems will have on the business? Are there systems that have less impact and can be designed as Tier 1 or Tier 2 to reduce cost?
- Redundancy vs. Cost—Designers need to understand the up-charge for implementing redundancy versus what the downtime of specific systems will cost the business (i.e., how much downtime can the business afford?)
- Agility/Speed to Deploy—How does the design criteria affect the ability and speed for the business to grow?
- Standardization & Industrialization—Think of the Data Center as a factory of compute power and consider standard configurations that allow for efficient scalability.
- Modularity—Avoid designing the entire space to today's requirements. Instead, think of the space in its entirety. Plan for tomorrow by separating larger spaces into smaller, more manageable spaces and sizing critical systems to allow for incremental growth and efficient operation without interruption.
- Total Cost of Ownership (TCO) vs. Build Costs—Carefully look at what the Data Center will cost to operate versus the capital expenditure to build it. Will certain energy saving strategies and solutions ultimately pay for themselves?
- Performance vs. TCO— How much computing and capacity are you actually getting for your operational costs? Will strategies and solutions ultimately cost more to operate than the savings they produce?
- Owner/Operator vs. Multi-tenant— As companies look to replace aging infrastructure and expand to meet customer demands, there is a growing trend to outsource all or some of the Data Center to a third party. Moving to a Multi-tenant Data Center (MTDC) avoids the high capital expenditure of upgrading infrastructure and supporting systems. When business growth and availability requirements demand 24/7/365 operation, businesses can leverage the MTDC's staff and expertise rather than facing the high operating expense of employing around-the-clock in-house expertise.
- Energy Efficiency –A given for any modern Data Center, energy efficiency must take into account several factors—from cooling and containment strategies to geographical location and how the Data Center is powered. Think free cooling as much as possible (70% or more) and look for geological advantages (i.e., rivers, lakes and geothermal) as a means to reduce carbon footprint and energy costs.
- Data Center Infrastructure Management (DCIM) —Gain a total holistic view of the Data Center by tracking, reporting and analyzing every aspect using DCIM solutions. With the asset management, operational management, reporting and analytics offered by DCIM, it is easy for Data Center managers to conduct modeling and future "what-if" scenarios before embarking on upgrades.
While the holistic approach to Data Center designs means literally questioning everything, Data Center managers that deploy holistic design principles will not only effectively reduce long-term costs and increase reliability, but they will also be better prepared for the future by "building it" rather than "fixing it."
Have questions or comments? Feel free to leave a comment below or you can schedule a call with one of our data center experts. I look forward to covering other Data Center trends in future posts.
Mike Salvador is a 28-year industry veteran, living the challenge of operating efficient data centers, optimizing the performance of network devices and delivering highly available, highly agile, low-risk data centers. Mike served as Belden’s technical solutions manager from 2012 to 2015.