We have been using it quite successfully for a bit now, and as we expand
our use cases we have discovered something odd.
The outer join works fine with up to 31 tables, but at 32, it appears
that the left most table is dropped from the join.
The job conf objects have the correct join statement in them when we
examine the failed jobs. The problem tracks 32 tables and not the which
tables happened to be the 32 tables.
Has anyone seen behavior like this?
--
Jason Venner
Attributor - Program the Web <http://www.attributor.com/>
Attributor is hiring Hadoop Wranglers and coding wizards, contact if
interested
