With around 8% of your energy costs coming from your data centre, it’s not surprising that optimizing efficiency is a priority for businesses.
With around 8% of your energy costs coming from your data centre, it’s not surprising that optimizing efficiency is a priority for businesses.
Over the years, business owners and managers have been taking steps to lower power expenditures but more needs to be done. Lowering power usage starts by identifying data center energy efficiency.
Taking steps like creating hot and cold aisles and integrating renewable energy can only do so much if your data centre has hidden energy hogs. Determining what’s using the most energy starts by measuring PUE (power usage effectiveness).
In this article, we’ll look at how PUE is calculated and how you can use the data to boost your data centre’s energy efficiency.
What Is PUE?
Power usage effectiveness or PUE refers to a metric used to measure the energy efficiency of a data centre. The total amount of power coming into the data centre is divided by the amount of energy needed to run the components.
You’ll quickly realize some components are energy hogs like servers and others like flash drives used for storage barely affect energy usage rates. Did you know that your servers may be responsible for using about 40% of your data center’s power?
So, why should you care about PUE beyond your energy usage costs? Industry regulations are changing. Data centres are responsible for a high percentage of the world’s energy usage. They’re also placing a strain on electrical grids. Don’t forget about greenhouse gas emissions.
Ideally, you want your data centre PUE score to be under 1.57, which is where most are currently averaging. However, before you can start working on lowering your PUE score, you need to know how to measure the metric.
How to Measure PUE in a Data Center
Okay, you can grab a pen, some paper, and a calculator and come up with your data centre’s PUE score manually.
While it’s possible, it’s also time-consuming and it’s easy to make a mistake in your calculations. If you come up with the wrong PUE score, you may not be able to identify the true energy hogs. Instead of identifying an outdated server, you may believe replacing interconnections will resolve your issue with high energy usage.
Thankfully, there’s an easier, faster, and more accurate way of determining PUE. You can use DCIM (data center infrastructure management) software. The software makes it easy to see how much power different devices are using. With this data, you can do the following:
- Track power consumption across the data center
- Analyze data to identify power usage trends
- Compare different energy usage optimization solutions
The software also demonstrates your data centre is taking steps to meet sustainability and regulatory goals.
Optimizing Data Center Efficiency with PUE Data
Chances are your data center’s PUE score is above the optimal goal of 1. Don’t worry; even industry giants like Google and Microsoft are also working on bringing down their PUE scores of 1.1 and 1.22 respectively.
The following steps can help lower PUE scores. Some are pretty easy to implement while others may take time and additional resources.
Eliminate Ghost Servers
Ghost, zombie, and comatose are all terms used to describe servers using energy without performing any functions. Maybe these are old servers no one has gotten around to disconnecting from the system. The servers may also be redundant. Whatever the reason for the ghost servers, they’re energy hogs.
Eliminating your ghost servers is one of the easiest ways to reduce power consumption and lower your PUE score. Best of all, you don’t need to send your IT staff around to check every server. Using DCIM software, you can identify and remove servers from the system. Yes, DCIM solutions can even identify and decommission virtual ghost servers.
Automate Data Center Monitoring and Management
You can do a lot with DCIM software to optimize data centre energy efficiency. Along with calculating your PUE score and identifying zombie servers, the software can also remotely monitor and manage data centres.
You can allow your teams to track data like power consumption trends and receive notifications if electricity usage spikes or a component fails. You can monitor and manage your cooling technologies and learn which ones are effective and which methods need a little tweaking.
Best of all, you can monitor and manage your data center’s energy usage remotely. So, if something goes wrong you can respond in real-time.
Improve Cooling Technologies
Data centres generate a lot of heat, and there’s no getting around this. Opening the doors and windows in a data center may give staff some fresh air but it’s not going to effectively keep your sensitive components cool.
You can take advantage of several tools to optimize cooling. Sensors can monitor everything from temperature and airflow to humidity levels. Creating hot and cold aisles can improve airflow efficiency. You may even want to invest in liquid cooling solutions.
Liquid cooling is efficient and reduces energy usage rates but the technology is an investment that may not fit every company’s budget.
Optimize Power Distribution
Optimizing power distribution in a data centre is something businesses can do in stages. If you’re working with a limited budget, this may be an ideal solution. Even making one change to how the power is distributed can make a difference.
For example, installing new circuit breakers is relatively inexpensive and you may not need to call in the professionals. Swapping out your old UPSs for more advanced models can also reduce energy usage while speeding up some computing processes. This is only a brief example of how you can optimize power distribution without going over budget.
You Can Lower Your Data Center’s PUE Score and Improve Energy Efficiency
If optimizing energy efficiency in your data centre is at the top of your to-do list, a good place to start is by calculating your PUE score. This makes it easier to identify your energy hogs and take steps to lower your score.
Don’t forget that lower PUE scores indicate greater energy efficiency. From removing ghost servers to automating some processes, you can make your data centre more efficient.
