Every year, it seems that the cost of data center downtime increases. At last count, the Ponemon Institute estimated that, on average, a single minute of downtime costs nearly $9,000, and shows no signs of coming down in the years ahead. 

More importantly, data is the lifeblood of modern business – and for that matter, of government, critical infrastructure, communication, health care and just about every other aspect of human society. Beyond the tangible monetary loss of downtime, data center disruptions in both the public and private sectors can have far-reaching, difficult-to-quantify consequences. 

In theory, the purpose of an uninterruptible power supply (UPS) is to act as the last line of defense between these data center disruptions and business as usual. However, they're more commonly cited as the top reason that a data center fails. Consider just a few of the following examples from the past two years:

1. In June of this year, a major airline in the U.K. experienced a UPS failure that "caused total chaos for tens of thousands of passengers, stranding them overseas, sending their luggage hither and thither, and grounding aircraft while frantic staff tried to sort out the madness," according to The Register. 
2. Across the pond, a major U.S. airline also experienced data center troubles earlier this year, resulting in the cancelation or delay of several thousand flights. The disruptions ensued after its secondary power source failed to resume operations.   
3. Last year, BT Group subscribers throughout the U.K. experienced internet connectivity issues after a colocation facility in London faced unplanned downtime after a UPS failure.

Bear in mind, these are only three of the most high-profile examples of UPS-induced downtime. Still, they get the point across: More can and must be done to prevent this prolific source of IT downtime. 

To that end, we have created a five-step graphic (see below) to help data center managers more effectively maintain their secondary power supplies. 

A few additional considerations

Before you dive in, we wanted to share a few additional notes pertaining specifically to steps one and three in the infographic:

For step number one

Don't forget about humidity and dew point. Temperature is perhaps the most important consideration where battery storage is concerned, but data center monitoring doesn't stop there. Most UPS units will have humidity thresholds as well, and those limits must be maintained. Not to mention, the last thing you want is to create a dank environment that will attract wire-chewing vermin (rats, squirrels, etc.). 

It also bears repeating that you should not treat your UPS room like a storage closet. Keep it free of dust and other airborne particulates. 

For step number three
Yes, having two PDUs per rack is crucial to making sure that power distribution will fail over instantaneously. However, that method only works if you correctly distribute your power load between your A/B feeds. Specifically, power drawn on each of those PDUs must not exceed 50 percent of the power strip's maximum load. Otherwise, if the primary power source fails, the secondary supply feed will be handling a load that exceeds 100 percent. 

The upside is that a reliable, well-made PDU should have no problem handling the maximum capacity determined by the manufacturer. But as a general rule, exceed that capacity, and even the most well-maintained, powerful UPS may not be able to distribute its load to all of your mission-critical equipment.

With those pointers in mind, here are the five simples steps that will help you prevent UPS-related downtime: