AnsweredAssumed Answered

How Each node is aware of Job tracker info.

Question asked by kumar on Mar 16, 2014
Latest reply on Mar 19, 2014 by gera

We are new to MapR and unable to understand how jobs are submitted to job tracker in MapR.

1) Assume I have written one simple Map/Reduce program and I am running it from cluster.

2) Now there is no Job tracker info provided in the core-site.xml or maprd-site.xml.
   (It is maprfs:/// in our case). So how this MR picks up job tracker info.

3) We checked warden.conf file and it seems from there it is starting the job tracker
   (But if we start warden on each node will it not start job tracker on each node)

4) If Zookeeper is locating the job tracker there is no where we configure Zookeepr URL.

We are confused why these basic things are not explained in MapR documentation.