You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@drill.apache.org by Gauri Srinivas <ga...@gmail.com> on 2018/07/13 00:06:34 UTC

Question on querying partitioned data

Hi,

I'm trying to use drill on data partitioned by day. As I understand it, the
performance tuning works by only querying the directories specified in the
where clause - in this case a query specifying the date.

Is there a performant way to query the partitioned data based on the
results of a join?

When I explain the query plan, it seems to be still querying all
directories rather than just the day from the output of the joined table.

Re: Question on querying partitioned data

Posted by Gautam Parai <gp...@mapr.com>.
What you are looking for is run-time partition pruning instead of
compile-time partition pruning.
AFAIK Drill currently supports compile-time partition pruning.

On Thu, Jul 12, 2018 at 5:06 PM, Gauri Srinivas <ga...@gmail.com>
wrote:

> Hi,
>
> I'm trying to use drill on data partitioned by day. As I understand it, the
> performance tuning works by only querying the directories specified in the
> where clause - in this case a query specifying the date.
>
> Is there a performant way to query the partitioned data based on the
> results of a join?
>
> When I explain the query plan, it seems to be still querying all
> directories rather than just the day from the output of the joined table.
>