Hadoop implementation over Peer-2-Peer

["jxtadoop@besnard.mobi" <jx...@hushmail.com>]

Hello,
I've just posted a work done for few months now:    
http://sourceforge.net/projects/jxtadoop/

My goal is to be able to run Hadoop on desktop computers which are
spread across the entreprise or across multiple locations over
Internet. For example, i plan to use it to share very large files
(kids videos ...) with other members of my familly spread across the
world instead of opening mutliple VPN to a focal point.
I know this is not the Hadoop use case since it only leverages the
HDFS side but this project helps to leverage unused storage capacity.
>From a usage standpoint, it is only possible to load/retrieve data
thru the DFSClient; hence no MapReduce (and i don't plan to do it).
BrgdsFranck
jxtadoop@besnard.mobi