You would specify all that by yourself.
And what do you mean by installing Map and Reduce on separate nodes?You
just have TaskTrackers which run continuously on each slave machine and
depending upon the location of the data block you are going to process maps
are started. Once the map phase is finished all the values associated with
a particular key are sent to the same machine for the reduce phase. It
could be both map+reduce on the same machine or on a different machine.
On Thu, Feb 14, 2013 at 11:53 PM, Henjarappa, Savitha <
[EMAIL PROTECTED]> wrote:
> My questions
> - If I have multiple Hadoop clusters, how do I make put which DataNode
> is configured to which NameNode? Similary Task Tracker to JobTracker?
> - Can I install Map and Reduce on separate Nodes? Is there any usecase
> to support this configuration.