AnsweredAssumed Answered

How to define pool via code for FairScheduler?

Question asked by chriscurtin on Jun 25, 2012
Latest reply on Jun 26, 2012 by chriscurtin
Hi,

We've enabled FairScheduler, restarted the JobTracker and can see the pools in the fairscheduler UI.

But all jobs submitted go into the 'hadoop' pool instead of the pool explicitly specified in the java code.

We're using Cascading so we set the mapred.job.pool.name to the pool we want when creating Flow, which is passed to the Hadoop job configuration.

Is there something different we need to do to set this for MapR?

We're running v. 1.2.3.12961.GA

Thanks,

Chris

Outcomes