Marathon Looks To Broaden Fault Tolerant Computing
- By Jeffrey Schwartz
- October 15, 2010
Marathon Technologies last week rolled out software aimed at broadening the availability of fault tolerant computing.
The company's new everRunMX software is designed to run on commodity Intel-architecture symmetric multiprocessing (SMP) and multi-core servers enabling them to function as fault tolerant systems. The software is intended for enterprises where continuous uptime is essential, according to company officials.
That might include media outlets, retailers and airlines, where any amount of downtime could be costly.
"In the past fault tolerant solutions were limited to a single processor, but now with our support for symmetric multiprocessing and multicore servers, we've broken that limitation" said Marathon president and CEO Jim Welch, in an interview. "Now the largest applications can be run, resource-hungry applications can be run, in our environment. That's a major, major breakthrough."
Marathon's new software lets customers stack workloads in a virtualized way to provide multiple workloads on that same platform, Welch said. "So for a small or medium-sized business, not only do they get the platform with everRun MX that gives them zero down time, but it also allows them to consolidate their servers onto fewer servers and save costs and power and management and resources all at the same time."
By establishing a fault tolerant environment, organizations can reduce the cost of maintaining environments that require high availability, he said "Traditionally, high availability required a lot of IT administrative staff to maintain, and they required specialized training," he added. That's because instead of creating two environments to manage, everRunMX provides one virtualized environment.
The software takes two off the shelf Intel-architecture servers and combines them in one environment. And through the software, it links them through an "availability manager," which provides the scalability for fault tolerant SMP. "This gives the power that up until now that has been only a dream for most customers," Welch said.
Marathon says it has 3,000 customers which include Boeing, Caesars, CNN, Delta Airlines, ESPN, General Dynamics, Harrah's, Proctor & Gamble and Walgreens.
"This is a proactive approach to ensure the highest level of system availability. They are saying it's not just high availability but mainframe-like fault tolerance," said Laura DiDio, principal of Information Technology Intelligence Consulting. "This is a proactive solution saying we're giving you SMP based fault tolerance in software. It's hardware agnostic and it's scalable; it's for today's virtual environments and what they're saying is it's better than the traditional restore, fail and restart."
With the release of everRunMX, Marathon is looking to bring the software to small and medium enterprises. The software has an entry level price of $10,000 for a pair of servers. That price covers all the cores and processors on the boxes and includes a year of maintenance.
Jeffrey Schwartz is editor of Redmond magazine and also covers cloud computing for Virtualization Review's Cloud Report. In addition, he writes the Channeling the Cloud column for Redmond Channel Partner. Follow him on Twitter @JeffreySchwartz.