queues, beeline/hs2 and tez

[Stephen Sprague]

hey guys (its me again!)

this is a simple one i think.   i'm trying to set the tez.queue.name via
the --hiveconf switch on beeline and it doesn't look to me it works.  the
question is... should it?


so this submits to the queue 'root.default' (which i don't want)

{code}
beeline \
    --hiveconf tez.queue.name=dwr.low \
    --hiveconf hive.execution.engine=tez \
    -u "jdbc:hive2://dwrdevnn1.sv2.trulia.com:10001/default;auth=noSasl
'spragues' nopwd org.apache.hive.jdbc.HiveDriver"  \
    --fastConnect=true  \
    -e "select count(*), date_key from omniture.hit_data_aws group by
date_key order by date_key"
{code}

while this submits to the queue 'dwr.low' (which i do want) but i don't
want to add that clause in red. :(

{code}
beeline \
    --hiveconf tez.queue.name=dwr.low \
    --hiveconf hive.execution.engine=tez \
    -u "jdbc:hive2://dwrdevnn1.sv2.trulia.com:10001/default;auth=noSasl
'spragues' nopwd org.apache.hive.jdbc.HiveDriver"  \
    --fastConnect=true  \
    -e "*set tez.queue.name <http://tez.queue.name>=dwr.low;* select
count(*), date_key from omniture.hit_data_aws group by date_key order by
date_key"
{code}

given i'm trying to write a general purpose sql wrapper i'd kinda like the
first one to work b/c the sql string used in the -e switch is coming from
the caller and i really don't want to inject my code into that if i can
help it.

anybody else run across this before? is there a trick?

thanks,
Stephen.
PS  when hive.execution.engine=mr, --hiveconf mapred.job.queue.name=dwr.low
works as advertised.