Hello,
i make install spark and jobserver in multi node mode, work fine,
now i look impala, it is working for hortonworks ?
because hive use map reduce and workflow execution, it take long time for get result, but impala doent use map reduce
and get 3 time faster with simple request seem to be good alternative, seem to choice of architecture and design mecanism.
it is apache version existing of impala ? i found only cloudera version
may be, is cloudera impala compatible with horton works, wich kind of version impala must be use for hadoop 2.6.0
i dont found it in component list for hortonworks.
KR
JP