Re: NEWBIE: excessive logging using parquet (from Hive)

[Sean Violante]

Thanks Brock for your assistance.  I am definitely not running locally.  I
see the reduce jobs all running on the hadoop clusters and see the
individual logs.  What I was trying to say is that the Hadoop logs (on
remote machines) show non zero bytes being written. Then after the hadoop
job has completed I see all the zero bytes chunks written on local console.
And it is this writing to the local console ( of 0B written) that is taking
~ day [the actual insertion takes on the order of 30 minutes].