Rack PDU Blog

In our previous post on 3-phase 208 power, we set the stage by describing some of the differences between it and single-phase power, including the fact that 120 + 120 does not equal 240 in the three-phase world. One thing that most of us know is that the rules for calculating capacity are different between the two. In this blog we will discuss the reason behind the difference and help build the case for the importance of a metered three-phase power strip.

Let’s be honest: Your data center is one of the most expensive components of your organization. For this reason, your key stakeholders may very well be considering outsourcing operations to a cost-effective colocation center at some point in the future if you don’t gain control over your spending. Why is your data center so expensive to operate? There are countless contributing factors. First and foremost, infrastructure is not cheap. Every year, a large portion of your budget is allocated toward hefty CAPEX and OPEX on equipment and cutting-edge technologies. Plus, there are workers’ salaries to consider, as well as costs related to heating, cooling and real estate.

Ask most people involved in data center operations to explain three-phase power to a data center PDU to you, and you’ll notice that they get a little squeamish. Except for your Principals of Electricity professor in college, most people have a difficult time explaining three-phase power and three phase PDU(s). Heck, most people have a tough time explaining how power works, period.  When it comes to rack mount power distribution, what is the difference between single phase and three phase power, you ask?

What happens once you get all of that power efficiently delivered to the rack? You guessed it. Server Technology is there to help distribute it by supporting your 240/415V power distribution needs through innovative rack PDUs.

On February 2, 2018, Server Technology will join, at least in spirit, the crowd gathered at Gobbler’s Knob outside Punxsutawney, PA. There, Groundhog Day celebrants will await the appearance of Punxsutawney Phil, the groundhog who will supposedly determine the length of winter by his ability to see his own shadow. We will also be joining countless others in their annual viewing of the Bill Murray classic, Groundhog Day. As movie fans know, Bill Murray plays a wretched reporter named Phil Connors, who finds himself stuck reliving the same day over and over again. Over the course of the film, which takes place over a ten-year period of real-time, Connors eventually redeems himself. Not only does he win the heart of Andie McDowell’s character Rita, but he also becomes a better person in the process. While much has been written about the spiritual and philosophical virtues of the classic comedy, the thing that strikes us is the idea that making improvements is a daily commitment.

Students may have left for the summer, but for data center professionals working for school systems and universities, the work is just getting started. Summer is a time to reflect on how the data center performed during the previous year, for identifying areas that need improvement and for making critical adjustments before September rolls back around. One issue that must be addressed during summer planning is data center power management. After all, power is a critical foundational element in the data center. But it’s also very expensive, and risky.

When considering implementing a power-dense data center, you face a number of challenges: cost, heat, load balancing, data center environmental monitoring and power monitoring, increased network traffic and system configuration management. With each of these challenges come serious, seemingly insurmountable obstacles, but with the proper planning and understanding, it can be done. As the Power Strategy Experts, Server Technology is here to help you successfully manage density and data center power.

It’s not unusual to be a little fuzzy on the differences between the two, but here it is at a high level: average rack power density is power utilization density that’s averaged over a period of time and space, and peak power density is a maximum power reading at a specific point in time.

In May and August of last year, British Airways was forced to cancel flights following a critical data center power outage. According to Network World, the trouble started when an engineer disconnected a critical power supply at a London-based data center. When the engineer tried to reconnect the system, it generated a large power surge that in turn caused the outage. Sources indicate that the power-up was handled “in an unplanned and uncontrolled fashion.” Altogether, 75,000 people were stranded at the airport following the incident. And the power surge also created a considerable amount of damage for the airline — to the tune of at least $112 million.

It’s not easy trying to decide which type of rack PDU to purchase for your data center. As we explained in a recent article, the decision ultimately comes down to your data center’s unique needs. Your appetite for risk, facility budget, and operating preferences will all play a big part in determining what type of rack PDU you should be using. In this article, we will focus on the switched PDU, which is a smart PDU that will let you toggle power on and off, on a per outlet bases, from a remote location. A switched PDU also allows you to centralize the management functions of your PDU.