Re: Possible regression (at least for me :D) in 4.2.M1-SNAPSHOT

Hi Lon.

Thanks for reporting this!
There is completely new SelectTranslator in 4.2 so there can be some
regressions. Will look at this one.

On Wed, Jan 9, 2019 at 11:49 PM Lon Varscsak <[email protected]> wrote:
>
> I'm experiencing crazy slow query times since I updated to this version.
> The only thing I see is that it's structuring the queries slightly
> differently.  For example:  4.1.M3-SNAPSHOT:
>
> SELECT DISTINCT t0.birthday, t0.contact_id, t0.contact_lead_source,
> t0.contact_status, t0.create_date, t0.glove_user, t0.graduation_date,
> t0.name_first, t0.name_last, t0.name_middle, t0.name_prefix,
> t0.name_salutation, t0.name_tag, t0.national_provider_id, t0.notes,
> t0.professional_designation, t0.specialty FROM customer.dbo.contact t0
> *JOIN* customer.dbo.contact_affiliations t1 ON (t0.contact_id =
> t1.contact_id) WHERE t1.company_id = ? AND t1.contact_id = ?
>
> and now:
>
> SELECT DISTINCT t0.birthday, t0.contact_id, t0.contact_lead_source,
> t0.contact_status, t0.create_date, t0.glove_user, t0.graduation_date,
> t0.name_first, t0.name_last, t0.name_middle, t0.name_prefix,
> t0.name_salutation, t0.name_tag, t0.national_provider_id, t0.notes,
> t0.professional_designation, t0.specialty FROM customer.dbo.contact t0 *LEFT
> JOIN* customer.dbo.contact_affiliations t1 ON t0.contact_id = t1.contact_id
> WHERE *( ( t1.company_id = ? ) AND ( t1.contact_id = ? ) )*
>
> The only thing I can think is that Sybase's (my DB) query optimizer isn't
> happy with the LEFT JOIN.  In DbPathProcessor it's always specifying a
> JoinType.LEFT_OUTER for relationship joins...this doesn't seem right to me,
> but I might be misunderstanding what's going on.
>
> Thoughts?
>
> -Lon



-- 
Best regards,
Nikita Timofeev