Thanks a lot Sunjay, Any more thoughts on this,
Im okie with if any alternative for Merge concept in Hive On Thu, Mar 6, 2014 at 2:02 AM, Subramanian, Sanjay (HQP) < [email protected]> wrote: > Hey Raj > > Maybe I am misunderstanding the question but u don't really have to do > anything fancy to merge > > ONE TIME > ======== > CREATE EXTERNAL TABLE employee ( > empno BIGINT, > ename STRING) ROW FORMAT DELIMITED FIELDS TERMINATED BY '\t' ; > > ALTER TABLE employee SET LOCATION > 'hdfs://path/to/dir/on/hdfs/containing/files' > > Or if u r using AMAZON EMR : > ALTER TABLE employee SET LOCATION > 's3://bucketname/path/to/subfolder/containing/files' > > Now if u keep putting files into this HDFS dir > 'hdfs://path/to/dir/on/hdfs/containing/files' > U should not have to do anything > > Thanks > > Warm Regards > > > Sanjay > > linkedin:http://www.linkedin.com/in/subramaniansanjay > > From: Raj hadoop <[email protected]> > Reply-To: "[email protected]" <[email protected]> > Date: Wednesday, March 5, 2014 at 4:16 AM > To: "[email protected]" <[email protected]> > Subject: Merge records in hive > > Hi, > > > > Help required to merge data in hive, > > > > Ex: > > Today file > > ------------- > > Empno ename > > 1 abc > > 2 def > > 3 ghi > > > > Tomorrow file > > ----------------- > > Empno ename > > 5 abcd > > 6 defg > > 7 ghij > > > > > > Reg: should not drop the hive table and then create it,what I actually > require is as shown in the example we have to merge the data, > > > > Thanks, > > Raj >
