Hadoop 1.2.1 - multinode cluster - Reducer phase hangs for Wordcount program?

Question

My question may sound redundant here but the solution to the earlier questions were all ad-hoc. few I have tried but no luck yet.

Acutally, I am working on hadoop-1.2

Answer 1

Got it fixed.. although, the same issue has multiple questions on the forums but the verified solution according to me is that hostname resolution for the any node in the cluster should be correct (moreover this issue doesnot depend upon the size of cluster).

Actually it is the issue with dns-lookup, ensure one make the below changes to resolve the above issue -

1. try printing hostname on each machine using '$ hostname'

2. check that the hostname printed for each machine is same as the entry made in master/slaves file for respective machine.

3. If it doesn't matches then rename the host by making changes in the /etc/hostname file and reboot the system.

Example :-

in /etc/hosts file (let's say on Master machine of hadoop cluster)

127.0.0.1 localhost

127.0.1.1 john-machine

#Hadoop cluster

172.50.88.21 HadoopMaster

172.50.88.22 HadoopSlave1

172.50.88.23 HadoopSlave2

then it's -> /etc/hostname file (on master machine) should contain the following entry (for the above issue to be resolved)

HadoopMaster

similarly verify the /etc/hostname files of the each slave node.