A new study warns that ignorant data center managers switching to high-density devices are facing a "capacity crisis" that may increase the number of costly server outages.
The Aperture Research Institute, a division of software company Aperture Technologies, surveyed over 100 enterprise data managers, representing over 600 data centers covering industries such as banking, data services, retail, insurance and telecommunications.
Almost 90 per cent of those surveyed said that their racks are already three-quarters full. This severe lack of real-estate is making high-density blade servers more appealing.
But with the greater complexity of blade servers and their intense demands of power and cooling, the risk of failure is also on the rise.
"Data centers are facing a time of crisis because of the increased demands on their physical resources and management," VP of marketing at Aperture, Steve Yellen said. "There's a gap between IT and data center facilities that's resulting in a rapid increase in high density equipment without thinking about the ability of a data center to reliably support that capacity. With these data centers stretching thinner and thinner, more and more instances of downtime and failure are likely to occur."
The survey found that over 57 per cent of of respondents cited human error as a leading cause of outages. Aperture worries that switching to long division while many are still counting on their fingers and toes is a recipe for a server outage sandwich.
Over 21 per cent of professionals surveyed are blissfully unaware of their maximum power density per rack. Over 18 percent didn't know what their average power density is. With more weight, heat and stress and power density per rack demanded by blades, Aperture argues this kilowatt apathy could lead to critical downtime.
You can take a gander at Aperture's findings yourself in PDF form.
Grain of salt
It should be noted that Aperture Technologies just so happens to have a data center infrastructure management suite available. Fancy that. ®