Limits and the practical usability of BSDs, a big data prospective

presented by Predrag Punosevac at BSDCan 2016

Abstract for talk:

The Auton Lab, part of Carnegie Mellon University's School of Computer Science, is a premier Statistical Data Mining group. We are interested in the underlying CS, Mathematics, Statistics and A.I. of detection and exploitation of patterns in data. We build practical large-scale deployments of very highly autonomous self-improving systems.

The Auton Lab computing infrastructure consists of about 40 large servers, 20 desktops and over 20 virtual instances, independently maintained from the scientific computing grid of the School of Computer Science. Three years ago I almost accidentally put on a system admin hat. Being a life long UNIX user and BSD hobbyist for the past 10 years I naturally tried to use BSDs whenever possible while rebuilding our internal network formerly ruled by penguins.

In this short talk I discuss limits and the practical usability of BSDs (Open, Free, and DF) and the obstacles of completely getting rid of penguins. Why is utilizing AWS further complicating my calculus?

About the Speaker: Predrag Punosevac

His principle research area is dynamical systems theory and applications. A dynamical system consists of a manifold whose points represent possible states of the system, "time" which may be continuous or discrete, and a time-evolution law. In some nonlinear systems similar initial states of the system lead to a very different future states. Such a phenomenon is known as chaos. As a consequence, the behavior of the system appears to be random despite the fact that system is deterministic.

Do you have anything to add? What did you find the most interesting? Maybe you have questions?

Please, join the discussion below!

comments powered by Disqus