How to save only non empty reducers' output in HDFS
It is possible - see the documentation section on "Lazy Output":
http://hadoop.apache.org/mapreduce/docs/current/mapred_tutorial.html#Lazy+Output+Creation
import org.apache.hadoop.mapreduce.lib.output.LazyOutputFormat;LazyOutputFormat.setOutputFormatClass(job, TextOutputFormat.class);