Companies require higher availability, but settle for less, survey says

June 10, 2010 - PRLog -- MAYNARD, Mass., USA and London, UK, June 10, 2010 – IT organisations are classifying more applications than ever as “business critical,” yet paradoxically risking hours or even days of downtime for those same applications that bring revenue through the door, according to a survey of more than 500 IT professionals.

ITIC’s “2010 Virtualization and High Availability Survey,” sponsored by Stratus Technologies, uncovered this apparent contradiction in responses from global IT professionals in industries including banking, government, manufacturing, high technology, retail, higher education, healthcare and engineering. The results come from the latest in a series of surveys on system uptime and related topics that Stratus and ITIC started in 2009 to track high-availability and continuous-availability computing trends.

The vast majority of IT groups are upgrading their application infrastructures and virtualising more of their data centres, according to the survey results. This is the logical time to implement highly reliable server infrastructures to support future growth, yet many IT groups are focusing on virtualisation and not the underlying infrastructure that supports it. Many IT managers incorrectly assume virtualised environments provide the same uptime as existing continuous availability systems. This, combined with the fact that most IT staffs don’t know how much an hour of downtime costs their companies, means that organisations are unwittingly in danger of higher risks and losses in the event of an outage.

“Virtualisation is a phenomenal technology, but the survey results suggest that people are relying on it for something it is not designed to do – provide continuous application uptime,” said ITIC Principal Laura DiDio. “There’s a cautionary tale emerging here about not losing sight of your fundamental IT needs when you weigh new technologies like virtualisation and cloud computing. A lot of companies are in for a big shock when they tally their losses from increased downtime and realise what a hit they’re taking in revenue, customer good will as well as the potential damage to their reputation.”

Down 10 Minutes per Week
The survey results indicate that IT managers assume the 99.9 percent uptime that a virtualised environment provides is enough for a critical application; this equates to core business operations being offline an average of ten minutes each week, or almost nine hours per year. Virtualisation’s “less is more” mantra with regards to server consolidation further complicates the uptime landscape. Slicing up one physical server’s resources into many virtual machines increases the impact of each server outage. Highly reliable infrastructures are even more vital in a virtualised environment to avoid losing many important applications with each failure.

The paradox in the survey findings coincides with the rapid rise of virtualisation in corporate data centres. Forty-three percent of the companies surveyed have virtualised at least half of their production servers, and 24 percent say 75 percent of their production servers are virtual. The overwhelming majority, 78 percent, have virtualised critical business applications. Another overwhelming majority – 86 percent – believe virtualisation provides sufficient uptime for critical applications.

“Rather than design for continuous availability, the majority of IT shops assume failure is inevitable and that recovery from failure is a reasonable design strategy,” said Roy Sanford Stratus CMO. “Accepting failure as ‘business as usual’ is reinforced daily by the large cloud services providers who make headlines for hours-long outages. The business and financial risks inherent in such a flawed IT methodology is costing individual enterprises hundreds of thousands of dollars, if not millions, every year.”

Andy Bailey, a UK based availability architect with Stratus Technologies, and prolific blogger, adds, “I see evidence of this everyday. The perception out there is that standard virtualisation technology comes complete with top level availability.  Sure virtualisation offers untold rewards in terms of cost, space and energy savings. The flip side of this though, is that it also offers untold levels of continuity risk. The good news is that by simply hosting on a server that doesn’t fail you can have your cake and eat it.”

Complete survey findings are available at the Stratus website.