Only for count(1) though. For others it still does 2mr. See hive-223 - it does what Qing is asking for. Still not committed - so can try out patch. 1mr with the option mentioned below. Will also do 1mr with hive.groupby.skewindata=false for non map-side aggregate as well.
________________________________ From: Zheng Shao [mailto:zsh...@gmail.com] Sent: Wednesday, February 18, 2009 8:14 PM To: hive-user@hadoop.apache.org Subject: Re: Is there a way to hint Hive the reduce key will be evenly distributed? Hi Qing, The easiest way to do that is to enable map side aggregation. In that case Hive will do one map-reduce job. hive> set hive.map.aggr=true; hive> explain select count(1) from mytable; Zheng On Wed, Feb 18, 2009 at 7:52 PM, Qing Yan <qing...@gmail.com<mailto:qing...@gmail.com>> wrote: So it does not have to run map reduce twice *every time*. Thank you! -- Yours, Zheng
