In past when using sqoop I've had luck using -Dmapred.job.queue.name=<QUEUE_NAME> in the sqoop command to specify a pool for the job to run in. However I've just been testing the hcatalog import and found that this value is getting ignored and the jobs are just running in root.<username>. I've also just tried exporting it into HADOOP_OPTS env variable but that seems to have had no effect either.
Has anyone had any luck setting an explicit pool for hcatalog import jobs? Thanks Josh
