High Performance Analytics on top of Hadoop

Welcome to 2015! We kick this year off with a special guest out from SAS software, Cary NC USA, Paul Kent (http://blogs.sas.com/content/sascom/author/paulkent/). Paul is a true veteran of data analytics platform development and is the Global VP of Big Data at SAS. Although this is short notice, I hope you can join us for an opportunity to take a peak under the hood of how SAS is leveraging the HDFS to support Big Data Analytics.

SAS is sponsoring the night at the Arthouse Hotel with drinks and pizza!

HDFS vs Posix File Semantics – almost the same but different: Paul Kent, VP Big Data at SAS

The distributed nature of the HDFS file system allows software some special opportunities to squeeze out extra performance, but it also imposes constraints on programmers used to the unix model of a file system. This talk describes the nature of HDFS, and how SAS has gone the extra mile in the name of performance gains. We discuss the tradeoff's between the convenience of the traditional API's versus the advantages gained by going one level deeper.