Nate Clyde, PE, is the Director of Data Center Design for Parallel Technologies.
To remain competitive in an increasingly competitive world, it is important to continually seek opportunities to boost operational efficiency, reduce expenses and improve the bottom line. Within every corner of business, improving efficiency is a never-ending journey.
As a significant capital investment, data centers are often under the microscope when it comes to improving performance. Within the industry, there is little doubt that data centers need to run as efficiently as possible to avoid tying up valuable (and often, unnecessary) company resources. In our business, we talk to data center managers every day and hear from them about their successes and frustrations when it comes to improving data center efficiency. It is evident from these conversations that there are five key areas where improvements can be made. Most importantly, these areas don’t require a lot of internal bandwidth but if done right, will go a long way towards an optimized data center.
- Data, Data, Data – For years, the American Society of Heating, Refrigerating and Air-Conditioning (ASHRAE) has recommended an increase in data center ambient temperatures. The reason for this is all about energy efficiency. Depending on the type of cooling system in the data center, each increase in ambient temperature (1 degree F) can result in a 2 to 5 percent decrease in annual energy utilization. Data center managers are reluctant to change the temperature because there is a lack of data that shows that the data center will still run effectively operating at the higher temperatures. Data centers that run at the ASHRAE recommended temperatures feel hot and I.T. people like data centers that feel cold. Unless there is data that shows that every device is operating within acceptable limits, the higher temperature simply don’t happen. However, there are now plenty of drop in solutions which provide operators the data they need to see that a higher temperature is not going to affect the data center and can actually result in energy savings. Plus, return of investment on these type of solutions is short, ranging from one to three years, making this approach a no brainer.
- Assets, Assets, Assets – Businesses move fast these days which requires I.T. infrastructure personnel to respond to constantly changing directions the business demands. Because of constantly shifting priorities, assets are often forgotten which can lead to “zombie servers”, which are assets that serve no purpose yet consume valuable data center space and power. To compound the problem, staff turnover is common and with little or no documentation left behind explaining the servers’ purpose, employees don’t know what to do with them or even know what they do! Too often no action is taken out of fear of causing an outage to customers leaving zombie servers sitting idle. The good news is that there are now a host of new solutions available that can ferret out zombie servers to ensure managers keep an accurate count on available capacity and avoid wasted resources.
- Back to Basics – Like everything, data centers have a growth cycle and unfortunately, many data centers are a mess, plain and simple. There are so many things that can pull a company in different directions but the key is to get back to basics in the data center. It may seem too basic but things like making sure abandoned cables are taken out and racks are sealed off from top to bottom can go a long way running an efficient data center. Without basic blocking and tackling in place, the rest of the data center will suffer.
- It’s All About Strategy – In response to ever changing business demands, I.T. infrastructure is also constantly changing. The need to build a bigger data center is an easy concept for leadership to get their minds around even when there is a company downsize. It’s important to make sure the data center strategy matches the company’s growth strategy. Otherwise, it results in unused capacity that doesn’t support the company goals. It is critical to have a strategic downsize strategy in place that ensures the data center continues to run effectively to meet the needs of the company.
- The Big Picture – We all know that I.T. assets are expensive and it can be tempting to consider purchases on the basis of lowest cost. However, it is important to look at the big picture as the cost to operate certain assets over their three to five year lifetime can sometimes exceed the cost of the device itself. For well-designed data centers, each kW of power draw that can be eliminated (generally through more efficient server/storage) counts twice for operations since the cooling system doesn’t have to account for that heat either, resulting in a double savings. For these reasons, it’s important to take the lifecycle of the equipment into account when making asset buying decisions. There are a number of vendor agnostic tools now available that can help end users make total cost of ownership comparisons between different I.T. equipment vendors. It may take more time but it’s certainly worth the extra effort in the long run.
While there are a lot more than five ways to optimize your data center, implementing these five steps will not only improve your data center operations but contribute to the overall performance of the business.
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.