https://community.cloudera.com/t5/Support-Questions/How-to-merge-multiple-part-files-while-creating-hive-ORC/m-p/176331#M138583 <P>Hi <A rel="user" href="https://community.cloudera.com/users/18374/saravananp1.html" nodeid="18374">@saravanan p</A></P><P>one way of doing it is modify the job to run it in single reducer so that the output will be a single file. </P><P>Use this property to change the reducer to one. set mapred.reduce.tasks=1;</P><P>By default the no of files inserted in a hive table depends on the size of a file, size of map job, size of reducer job. Based on the size no of files inserted in a hive table varies.</P><P>max(mapreduce.input.fileinputformat.split.minsize, min(mapreduce.input.fileinputformat.split.maxsize, dfs.block.size))</P><P> If you have reducers running, then you should also look at</P><P>hive.exec.max.created.files, mapred.reduce.tasks, hive.exec.reducers.bytes.per.reducer</P><P>Hope it helps!</P> Wed, 17 May 2017 17:34:53 GMT balavignesh_nag 2017-05-17T17:34:53Z