[rules-dev] Decision Tables with O(10000) rows

Michael Anstis michael.anstis at gmail.com
Tue Jun 19 06:47:39 EDT 2012


https://github.com/droolsjbpm/drools/commit/40721a87a7794af697223bb76d8d16823d0beeec

On 19 June 2012 10:49, Wolfgang Laun <wolfgang.laun at gmail.com> wrote:

> On 19/06/2012, Michael Anstis <michael.anstis at gmail.com> wrote:
> > Agreed.
> >
> > I'll add something.
> >
>
> Nice, thank you.
>
> > I've raised a JIRA to capture a need to shift to a different Excel
> library
> > too: https://issues.jboss.org/browse/JBRULES-3549
> >
>
> The times for (de)serialization indicate that it isn'n just Excel that
> doesn't. But big is big...
>
> -W
>
>
> > With kind regards,
> >
> > Mike
> >
> > On 19 June 2012 10:18, Wolfgang Laun <wolfgang.laun at gmail.com> wrote:
> >
> >> A recent post on the users list has led me to investigate DTs with
> >> O(10000) rows. While the sample XLS with 26,400 rows and 6 condition
> >> and 1 action columns was still manageable, a (my) system may exhibit
> >> severe memory problems when the number of rows is increased.
> >>
> >> Test indicate that ~25,000 rows is just about manageable if you don't
> >> need lightning startup for your application - serialization doesn't
> >> help here.
> >>
> >> It would appear that Drools users are (not intentionally) led to
> >> believe that Drools' performance is immune against the usual
> >> detriments that arise from "large numbers".
> >>
> >> I think that Section 2.4., Decision Tables in Spreadsheets, should
> >> contain a warning.
> >>
> >> -W
> >> _______________________________________________
> >> rules-dev mailing list
> >> rules-dev at lists.jboss.org
> >> https://lists.jboss.org/mailman/listinfo/rules-dev
> >>
> >
> _______________________________________________
> rules-dev mailing list
> rules-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/rules-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.jboss.org/pipermail/rules-dev/attachments/20120619/8221f8c0/attachment.html 


More information about the rules-dev mailing list