By nature, data centers are tooled for maximum reliability, and that means optimal performance, power and capabilities. It’s a configuration that breeds excess, albeit necessary excess, but all the same.
Anything in excess is essentially waste — wasted resources, wasted capital, and wasted potential. The wastefulness is exacerbated even more when finite resources are factored into the equation, such as power or additional space.
What this means is that any one organization or service provider can easily outgrow its current configuration, particularly the assets available within the local data center. It encourages constant growth and expansion, which is something that might not be necessary if and when efficient operations are applied.
As a way to cut down on some of that excess, it makes sense to improve efficiencies across the board, and there are certainly ways to do that in the modern data center.
It’s no secret that data center equipment and servers put off a lot of heat, which means a large portion of expenditures come from cooling and air conditioning. The equipment must remain at a safe temperature, which calls for proper ventilation and cooling in the housing room.
That power consumption can be lessened by optimizing not just the cooling operations, but also the space where the equipment is housed. Proper insulation, for example, can help maintain temperatures within the room. Strategic layouts of equipment and streamlined airflow can also improve cooling efficiencies.
Some additional measures managers can take include the following:
Because cooling is so important, most data center managers are unwilling to experiment to find more efficient temperatures. In reality, lowering maintained temps by even a couple degrees can save hundreds — if not thousands. It cuts down on power consumption and has a minimal effect on performance.
Spend some time monitoring temperature changes to find a level that works, but also allows for a boost in savings.
Even newer or refreshed configurations tend to waste power and resources when demand is low. It makes more sense to match up the server capacity — or at least active hardware — to meet demands in real-time.
With proper planning and the help of monitoring and management tools, it is possible to match-up these two elements to create a more streamlined system.
Thanks to a fast-moving business with constantly shifting operations, staff and processes, certain assets are either overlooked or forgotten now and then. This leads to something called a zombie server, a system that is no longer used yet remains powered on and consuming energy. Research shows that 25 percent of physical servers and 30 percent of virtual servers are comatose, or zombies.
Generally, they are not shut down because there’s no paper trail about what they contain or what they’re used for, meaning managers are afraid to hit the killswitch.
To deal with this problem properly, everything must be documented appropriately, and monitoring tools must be put in place to offer direct oversight as to what servers or configurations are necessary — what’s mission-critical?
Before server virtualization was possible, it was critical to outfit additional space with more servers to keep up with power and load demands. That is no longer the case — in fact, it may be more beneficial to downsize favoring optimized use of the space available instead.
Focusing on a modular design that can be scaled up or down to meet company needs is a great idea for maintaining proper efficiency levels.
In either case, excess space can balloon costs — especially when it must be factored into cooling and air control.
Every business or enterprise has a supplier of some kind. Data centers generally source energy from one or many suppliers, which is a contention point for elevated costs. By striking up more beneficial relationships or partnerships, those costs can be mitigated.
Furthermore, simply finding a good energy supplier can net you more savings through good communication. The said supplier can help you improve energy usage, source fuel or energy at cost, and even cut down on the time invested in such matters. The ideal energy supplier will help you better manage time because you can focus on other, more important matters like budgets, inventory, negotiations and more.
By optimizing certain operations and processes — temperature control and cooling primarily — it’s possible to improve the efficiency of a data center and reap cost savings in the process. In some circles, this is almost unheard of, as data centers tend to be a powerhouse of consumption and excess. They demand huge loads of energy and must remain online at all times of the day and night, which requires incredible levels of reliability and performance.
Still, there’s no reason not to at least attempt to mitigate some of that excess. By following the six tips discussed here, it's more possible than you may think.
Kayla Matthews writes about data centers and big data for several industry publications, including The Data Center Journal, Data Center Frontier and insideBIGDATA. To read more posts from Kayla, you can follower her personal tech blog at ProductivityBytes.com.