Dr. Konstantin Boudnik, WANdisco, Hadoop Summit 2013 with John Furrier and Dave Vellante Dr. Konstantin Boudnik is the Director of Engineering and Big Data Distribution for WANdisco. In this interview, he joins John Furrier of SiliconAngle and Wikion's Dave Vellante, inside theCUBE, at Hadoop Summit 2013. Traditional high availability architecture in HDFS allows users to copy over edit logs, but usually comes with downtime (ranging anywhere from 1 to 5 minutes or half an hour). Boudnik states that for users that absolutely need 100% availability, WANdisco is the sole option. Boudnik makes a bold claim, according to Furrier, in guaranteeing 100% uptime with WANdisco. Boudnik acknowledges that the solution will not prevent switches from getting blown up in Iraq, for example. However, he explains, "When it comes to availability of nanenodes in HDFS, every single client of yours can work with one of the masters. It actually depends on your use case how many namenodes you want to have around." WANdisco allows every namenode to have its own main copy of the metadata. As such, Boudnik notes, "Instead of dealing with the issues once they happen, we guarantee that the issues don't happen." The discussants also review economics. Boudnik explains that most companies are not as expansive as Google and Yahoo!. WANdisco is a reasonably price solution for businesses that can't afford to hire staff as large as industry leaders. BigTop is another open source Apache product with a framework that guarantees and allows users to build software stacks with pre-set software characteristics. Boudnik suggests this product is ideal for users that want to build RPM packages allowing customers to define, build, deploy and test a stack. Vellante asks about WANdisco's corporate culture asking how the company manages to attract its pool of talent. Having over 20 years of experience in the field and 14 U.S. patents, Boudnik believes WANdisco is at the absolute "bleeding edge" of technology in the field, which excites experts like him. @thecube #hadoopsummit