Not knowing much about internal working of hadoop, I was thinking maybe it is better to have smaller storage per node to better distribute the load between more nodes. Is this a correct way of thinking?
-----Original Message----- From: Edward Capriolo [mailto:[email protected]] Sent: Monday, November 16, 2009 11:50 AM To: [email protected] Subject: Re: hive web interface problem On Mon, Nov 16, 2009 at 11:07 AM, Massoud Mazar <[email protected]> wrote: > This mailing list has been a great source of information for me. Thanks we try :) > > Are there guidelines to select hardware configuration for hive nodes? > None in particular. Hive nodes are just regular DataNode/TaskTrackers. For reference, the hive code only had to be installed on the node launching the Job.
