drill-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From François Méthot <fmetho...@gmail.com>
Subject Re: Question regarding drill Query
Date Sat, 24 Sep 2016 02:56:24 GMT
If possible, I strongly suggest that you store all your ids in a csv file,
make the file available to the drillbit then you do something like

   select * from ... as tbl1 where tbl1.id in (select columns[0] as id from
dfs.`/millions_ids.csv.`)

If your ids are integers, you are in luck because the in statement on
integers is surprisingly fast.





On Thu, Sep 22, 2016 at 3:37 PM, Gautam Parai <gparai@maprtech.com> wrote:

> It might be helpful to split these IDs into ranges. Is it possible to do
> so?
>
> On Thu, Sep 22, 2016 at 12:02 PM, Ranjit Shukla <
> ranjitshukla1234@gmail.com>
> wrote:
>
> > Hi
> >
> > Question:
> >
> > *How can I configure drill to accept large query length?*
> >  I am facing this difficulty while querying with drill. My query is a bit
> > lengthy, something like IN operator with few million IDs. When I send
> this
> > query to drill, it does not respond after that. I am having  to restart
> the
> > drill server after that. Even though I have provided 10GB heap size in
> > drill-env.sh, its not returning the result.
> >
> >
> > Thanks and regards
> >
> > *Ranjit Shukla*
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message