Enhanced Power Reliability for Data Centers
January 31st, 2012 By: Industry Perspectives
Bhavesh Patel is the director of Marketing for ASCO Power Technologies, a business of Emerson Network Power.BHAVESH PATEL
Emerson Network Power’s ASCO business
This year I’ve noticed a few big-name data centers lost power for significant amounts of time due to preventable power reliability measures that shouldn’t ever be overlooked. Data center power reliability can be enhanced by employing three measures that work hand in hand: testing, maintenance and modernization.
Maintenance Is Critical
Maintaining a data center is a complex task that requires well-trained professionals to keep operations running smoothly. As complex as maintaining a data center is, the basic maintenance principles parallel those for regular car maintenance. And we can all identify with the need for car care and repair.
For instance, 99.99 percent of the time, your car will start when you insert the key into the ignition and turn. Cars are so reliable that we almost always take them for granted – until the day the engine turns over, but fails to start. Then, frustration sets in. The same could be said about data center power reliability when it’s operational and when it fails.
For most of us, cars are our lifeline to work, school, social events and everything in-between. In many ways, vehicles are a necessity to daily life. So, too, are data centers. They are the lifelines to the businesses we work for, the social media we use and the financial institutions we rely upon. We maintain our vehicles to ensure they operate reliably and should do the same for data centers.
Power Equipment Requires Maintenance and Testing
So when I read that transfer switches fail when utility power is lost, I wonder about the maintenance program for those switches and the on-site power systems. Maintaining and testing on-site power systems are integral elements of high power reliability. Maintenance programs cost pennies on the dollar compared to revenue lost when a data center’s backup power system fails. Lost revenue caused by interrupted business continuity either to the data center’s owner or its customers, if it’s a hosting service, tallies into the millions annually.
Emerson recently released a white paper titled, “Understanding the Cost of Data Center Downtime,” and the figures listed are sobering.
Data center downtime can be loosely compared to car trouble. Maintenance and testing to a data center is like changing the oil in your car. It’s an easy task and relatively inexpensive. But procrastinating is also quite convenient to make maintenance less of a priority. The car and data center are running fine now and will continue to do so for the next few days. Falling into this line of thinking is easy. After a while, the engine could seize up, costing thousands to repair or replace as opposed to paying $30 to $70 for an oil change. Now, think of the costs associated with on-site power components and downtime.
Without a regular maintenance and testing program, it doesn’t matter if your power system is rated Tier I or Tier IV, the likelihood of failure increases. Often times, a corporation will talk about the level of redundancy within the power distribution system. Without regular testing and maintenance, all that equipment may still start when required. But, is “may” good enough? Cars also are redundant in that most carry spare tires. But what if the spare has no air? Can you guarantee that your spare tire will be useable? When was the last time your spare tire was checked?
But maintenance can only go so far in ensuring the maximum power reliability. Legacy equipment, or equipment more than 20-years old, may be working properly without any major glitches. This equipment, however, is more prone to failure than newer equipment with the same maintenance program in place.
Buy New or Keep the Old?
In today’s rough economy, many corporations don’t have the budget to replace older gear with new equipment. There’s a way to ensure power reliability within a tight budget. It’s modernizing equipment. Modernizing, or Mods, means upgrading legacy equipment with new components. Mods provides three significant benefits for IT and facility managers: The process costs less than replacing the entire system, takes less time and minimizes disruption to business continuity.
Typically, modernizing existing equipment will save half to two-thirds of the cost of replacing the system. When capital expense budgets are tight, modernizing means there’s more money for refreshing compute equipment. In all, modernization to a data center is the equivalent of giving your car a big-time tune up to enhance power reliability for the long haul.
Power Gear Reliability
Modernizing, maintaining and testing transfer switches and other on-site power gear are integral to ensuring power reliability at a data center’s design level, whether its four, five or more nines.
Unfortunately, power reliability can create complacency. When the news stories hit the press and the Internet about power failures and downtime, most people think to themselves that losing power will never happen to them. The cruel reality is that there is never a good time to lose power or for the car to fail. Never.
We take care of our cars because they are a necessity for most of us. We need to take better care of our data centers, too, and that can be achieved by enhancing power reliability through modernization, maintenance and testing.
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.