I thought hash joins were fixed in ingres 26sp2 patches. I turned off opf_hash_join at all of my sites in sp1. Have not been game to turn them on.
-----Original Message----- From: Jose L Levieux [mailto:[EMAIL PROTECTED] Sent: Saturday, 25 June 2005 6:59 AM To: [email protected] Subject: Re[2]: [Users] Bad result sql query Thank's Julian It's like yuo say I change this setting and now it's work ok. We appreciated your help. Regards José Levieux Friday, June 24, 2005, 5:57:48 AM, you wrote: > Please Try This. > Shutdown your Ingres R3 installation. (ingstop) > Go into CBF. > Choose DBMS Server and press the configure button. > Set opf_hash_join to off. > Come out of CBF and (ingstart) > Try again, this solved many many of our sql problems. > Regards Julian Sedgewick. >>>> Jose L Levieux <[EMAIL PROTECTED]> 23/06/2005 21:32:50 >>> > I notice that if I select the column that I use to join in the > subselect the query work fine. > select 'c' as documento, > c.serie , c.numero, c.fecha, > o.codigo_unico, o.codigo, o.color, o.talle,c.sucursal, > p.precioventa, o.venta, money(0) as preciolista > from contados c, detallecontados d, codigos o, codigosprecio p > where c.fecha between '010105' and '033105' and > c.serie= d.serie and > c.numero = d.numero and > d.codigo_unico = o.codigo_unico and > d.codigo_unico = p.codigo_unico and > not exists (select z.sucursal > from zonasucursal z > where c.sucursal = z.sucursal) and > o.codigo != 65745 ; > (54056 rows) > End of Request > What append this ? Why I have to display > this column (I test this using the isql utility) > Saturday, June 18, 2005, 12:52:06 PM, you wrote: >> Hi , >> We are testing ingres R3 , before migrate from ingres 2.0 and we > found >> something strange. When whe change the columns that select the > result >> change , look this : >> select 'c' as documento, >> c.serie , c.numero, c.fecha, >> o.codigo_unico, o.codigo, o.color, o.talle, >> p.precioventa, o.venta, money(0) as preciolista >> from contados c, detallecontados d, codigos o, codigosprecio p >> where c.fecha between '010105' and '033105' and >> c.serie= d.serie and >> c.numero = d.numero and >> d.codigo_unico = o.codigo_unico and >> d.codigo_unico = p.codigo_unico and >> not exists (select z.sucursal >> from zonasucursal z >> where c.sucursal = z.sucursal) and >> o.codigo != 65745 ; >> (9 rows) >> End of Request >> > ----------------------------------------------------------------------- >> select c.* >> from contados c, detallecontados d, codigos o, codigosprecio p >> where c.fecha between '010105' and '033105' and >> c.serie= d.serie and >> c.numero = d.numero and >> d.codigo_unico = o.codigo_unico and >> d.codigo_unico = p.codigo_unico and >> not exists (select z.sucursal >> from zonasucursal z >> where c.sucursal = z.sucursal) and >> o.codigo != 65745 ; >> (54056 rows) >> End of Request >> > ------------------------------------------------------------------------- >> Look that the predicate it's the same in both requests. >> In ingres 2.0 both sqls works correctly. >> Jose L Levieux > _______________________________________________ > Users mailing list > [email protected] > http://ingres.ca.com/mailman/listinfo/users > _____________________________________________________________________ > This e-mail has been scanned for viruses by MCI's Internet Managed > Scanning Services - powered by MessageLabs. For further information > visit http://www.mci.com > _____________________________________________________________________ > This e-mail has been scanned for viruses by MCI's Internet > Managed Scanning Services - powered by MessageLabs. For further > information visit http://www.mci.com -- Best regards, Jose mailto:[EMAIL PROTECTED] _______________________________________________ Users mailing list [email protected] http://ingres.ca.com/mailman/listinfo/users _______________________________________________ Users mailing list [email protected] http://ingres.ca.com/mailman/listinfo/users
