|| at Jun 5, 2009 at 9:40 am
I have setup machines just to act as HADOOP clients which are not part
of the actual cluster (master/slave config). The only thing is that
these machines acting as hadoop clients were all internal in our network
and I have not tested with remote machines outside our internal LAN. My
assumption is if the access privilages are set right from the remote
machines (as clients) pointing to the namenode in the cluster you could
probably place data into HDFS without issues.
I have a 7 node cluster working as of now. I want to place the data into
HDFS, from a machine which is not a part of the hadoop cluster. How can I do
that.? It's in a way, a remote machine.
Will I have to use RPC mechanism or simply I can use FileSystem api and do
some kind of coding and make it work?