If you point your poolname property to mapred.job.queue.name, then you can leverage the Per-Queue ACLs described at http://hadoop.apache.org/docs/stable/mapred_tutorial.html#Job+Authorization to do user/group based control.
In addition, depending on the version/distribution of Apache Hadoop in use, you can set mapred.fairscheduler.allow.undeclared.pools to false in mapred-site.xml to disallow dynamic pool names (to box users to use specific poolnames and not get away with new ones). On Sun, Oct 14, 2012 at 6:03 AM, Patai Sangbutsarakum <[email protected]> wrote: > Is that anyway to control who can submit job to a pool.? > > Eg. Pool1, can run jobs submitted from any users except userx. > > Userx can submit jobs to poolx only. Can't submit to pool1. > > Hope this make sense. > Patai -- Harsh J
