So as Insert into isn't really an option, I would love some feed back on
options with Drill for ETL. I understand Drill is very good with raw text,
at the same time, for a highly used table, the advantages of taking text
data and "final" storing it as Parquet would be handy. I was thinking
through
You are probably on the right track to create a workspace for querying one or 2
levels up on the directory structure, and then use CTAS on subdirectories for
incremental data loads.
—Andries
> On Nov 2, 2015, at 5:29 AM, John Omernik wrote:
>
> So as Insert into isn't really an option, I wou
What are the thoughts for combining all the data in the individual
directories? Is there an option to replace like in Hive?
On Mon, Nov 2, 2015 at 9:27 AM, Andries Engelbrecht <
aengelbre...@maprtech.com> wrote:
> You are probably on the right track to create a workspace for querying one
> or