If I wrote "return" in reducer class like
public class reducerClass{
public void reduce{
if(counter <= 0)
return;
}
}
The program has to go through all the mapper keys in order to return.
I even want to stop all mapreduce process before the mapper class does the
sorting and shuffling.
Is there anything that I can call to exit mapreduce and return back to the main
method?
Thanks
Wei
-----Original Message-----
From: Peng, Wei [mailto:[email protected]]
Sent: Fri 12/17/2010 2:14 PM
To: [email protected]
Subject: RE: Please help with hadoop configuration parameter set and get
My question is how to exit hadoop reducer.
Can I just write "return;" in reducer?
Wei
-----Original Message-----
From: Aman [mailto:[email protected]]
Sent: Fri 12/17/2010 1:24 PM
To: [email protected]
Subject: RE: Please help with hadoop configuration parameter set and get
Set the counter in the mapper and in the reducer code, check for this
counter. If counter is set then run the reducer if not then exit without
doing anything.
Once there is no new node is found, I would like to exit mapper or reducer
process.
For example, if no new nodes are found in the mapper class, the setup of
reducer will stop reducer work (or the cleanup of mapper will stop the
reducer work).
--
View this message in context:
http://lucene.472066.n3.nabble.com/Please-help-with-hadoop-configuration-parameter-set-and-get-tp2103303p2107388.html
Sent from the Hadoop lucene-users mailing list archive at Nabble.com.
