Hi,
-- I merge data using my M/R job
-- For better compression and code maintainence, I want to replace my M/R job
with a SORT BY and DISTRIBUTE BY hive query
-- I also want to know the number of records ( hadoop counters ). I could do
that in my M/R job. Its a MUST.
-- I dont want to write second M/R job to get counts
I have following questions:
-- Do I have access to hadoop counters if I am using a hive query?
-- Looked into HIVE code base and looks like there are no public
interfaces
for me to either get JobClient or counters
-- ExecDriver.java is the class that deals with it and this class/methods
are internal to HIVE.
Please suggest a way to accomplish the above tasks.
Hoping to get a response soon!
Regards,
Gaurav Jain
