Hi Owen- On Sat, Jul 16, 2011 at 09:14:59AM -0700, Owen O'Malley wrote: > The scientists at CERN use HDFS for storing their large data sets and > don't use MapReduce at all. (I believe most of their applications are > MPI.) At Yahoo, some users run non-MapReduce in the shared MapReduce > cluster.
Various groups supporting LHC physics at CERN do have about a dozen petabyte-sized HDFS clusters. Our workflows are highly parallel, so we mostly use traditional schedulers like Condor[0], PBS and LSF. Thanks! [0] http://www.cs.wisc.edu/condor -- Will Maier - UW High Energy Physics cel: 608.438.6162 tel: 608.263.9692 web: http://www.hep.wisc.edu/~wcmaier/
