Have you seen the book Google's PageRank and Beyond? :) >> they might be using MapReduce ...
I don't think Map/Reduce is a advanced parallel computing model, but i'm agree with you. Have you seen the Hama proposal? (http://wiki.apache.org/incubator/HamaProposal) I'll presentation ideas about Hama matrix features and some algorithm strategies at Monday April 21. and Perhaps we could share these resources by creating the documents on the web as detail as possible. Thanks. On Wed, Apr 16, 2008 at 3:17 AM, Chaman Singh Verma <[EMAIL PROTECTED]> wrote: > > Hello, > > After googling for many days, I couldn't get one answer from many of the > published reports on > Ranking algorithm done by Google. Since Google uses GFS for fault tolerance > purposes, what > communication libraries they might be using to solve such a large matrix ? I > presume that standard > Message Passing (MPI) may not be suitable for this purpose(Fault tolerance > issue). I also presume > that they might be using MapReduce to create sparse matrix on distributed > nodes. > > Suppose we implement ranking algorithm on top of Hadoop, what could be the > best way/best > distributed algorithm/library etc ? Some hint from Google implementation > could be extremely useful. > > With regards. > csv > > ----- > Chaman Singh Verma > Poona, India > -- > View this message in context: > http://www.nabble.com/EigenValue-Calculations%2C-Hadoop-and-MPI.-tp16703423p16703423.html > Sent from the Hadoop core-user mailing list archive at Nabble.com. > > -- B. Regards, Edward J. Yoon
