Re: [SQL] Extremely Low performance with ODBC

2008-05-28 Thread Bruce Momjian
Sebastian Rychter wrote: > Thanks. Hiroshi sent me a possible solution (which might be useful for > others going through similar situations), which is to toggle off the > "Disable Genetic Optimizer" from the ODBC driver configuration. Uh, why is ODBC modifyingin the 'geqo' variable? -- Bruce M

Re: [SQL] Extremely Low performance with ODBC

2008-05-28 Thread Sebastian Rychter
e de Steve Midgley Enviado el: MiƩrcoles, 28 de Mayo de 2008 03:29 a.m. Para: [EMAIL PROTECTED]; pgsql-sql@postgresql.org CC: [EMAIL PROTECTED] Asunto: Re: [SQL] Extremely Low performance with ODBC At 09:20 PM 5/27/2008, [EMAIL PROTECTED] wrote: >Date: Tue, 27 May 2008 09:29:56 -0700 >From: &

Re: [SQL] Extremely Low performance with ODBC

2008-05-27 Thread Steve Midgley
At 09:20 PM 5/27/2008, [EMAIL PROTECTED] wrote: Date: Tue, 27 May 2008 09:29:56 -0700 From: "Richard Broersma" <[EMAIL PROTECTED]> To: "Sebastian Rychter" <[EMAIL PROTECTED]> Cc: pgsql-sql@postgresql.org Subject: Re: Extremely Low performance with ODBC Message-ID: <[EMAIL PROTECTED]> On Mon, M

Re: [SQL] Extremely Low performance with ODBC

2008-05-27 Thread Steve Midgley
At 09:20 PM 5/27/2008, [EMAIL PROTECTED] wrote: Date: Tue, 27 May 2008 09:29:56 -0700 From: "Richard Broersma" <[EMAIL PROTECTED]> To: "Sebastian Rychter" <[EMAIL PROTECTED]> Cc: pgsql-sql@postgresql.org Subject: Re: Extremely Low performance with ODBC Message-ID: <[EMAIL PROTECTED]> On Mon, M

Re: [SQL] Extremely Low performance with ODBC

2008-05-27 Thread Richard Broersma
On Mon, May 26, 2008 at 9:34 PM, Sebastian Rychter <[EMAIL PROTECTED]> wrote: > I'll keep on looking for any other differences on different logs.. I think I > already searched all the Postgresql forums I know and no one ever reported > something like that before. There is one other thing that you

Re: [SQL] Extremely Low performance with ODBC

2008-05-26 Thread Sebastian Rychter
thing like that before. Thanks once again, Sebastian -Mensaje original- De: Richard Broersma [mailto:[EMAIL PROTECTED] Enviado el: Martes, 27 de Mayo de 2008 01:14 a.m. Para: Sebastian Rychter CC: pgsql-sql@postgresql.org Asunto: Re: [SQL] Extremely Low performance with ODBC On Mon, M

Re: [SQL] Extremely Low performance with ODBC

2008-05-26 Thread Richard Broersma
On Mon, May 26, 2008 at 8:26 PM, Sebastian Rychter <[EMAIL PROTECTED]> wrote: > 2008-05-27 02:55:10 GMT LOG: duration: 112157.000 ms statement: SELECT > 2008-05-27 03:07:29 GMT LOG: duration: 250.000 ms statement: SELECT Wow, that is a big difference. I reformated your querys and diff'ed the

Re: [SQL] Extremely Low performance with ODBC

2008-05-26 Thread Sebastian Rychter
Title_ID" AND "Patient"."City_ID"=PatientCity."City_ID" AND "Patient"."Province_ID"=PatientProvince."Province_ID" AND "Patient"."Country_ID"=PatientCountry.

Re: [SQL] Extremely Low performance with ODBC

2008-05-23 Thread Richard Broersma
On Fri, May 23, 2008 at 2:45 PM, Sebastian Rychter <[EMAIL PROTECTED]> wrote: > Thanks for your response. I did that in the first place and found that both > explain reports where quite similar. I've got both reports: From pgadmin it > took 7288 ms, and from VFP took 6446 ms (of course, that's the

Re: [SQL] Extremely Low performance with ODBC

2008-05-23 Thread Sebastian Rychter
ows=1 loops=1) Hash Cond: ("PatientDoctorProvince"."Province_ID" = "Patient"."DoctorProvince_ID") -> Seq Scan on "Province" "PatientDoctorProvince" (cost=0.00..1.78 rows=78 width=4) (actual time=0.005..0.150 rows=78 loops=1) ->

Re: [SQL] Extremely Low performance with ODBC

2008-05-23 Thread Richard Broersma
On Fri, May 23, 2008 at 2:07 PM, Craig Ringer <[EMAIL PROTECTED]> wrote: > Can it be done? Or is the DB server not capable of generating explain output > (say to a log) and also returning a resultset? It can be done ~kind-of~. To me EXPLAIN [ ANALYSE ] is similar to any other SELECT statement qu

Re: [SQL] Extremely Low performance with ODBC

2008-05-23 Thread Craig Ringer
Richard Broersma wrote: On Fri, May 23, 2008 at 9:47 AM, Sebastian Rychter <[EMAIL PROTECTED]> wrote: Thanks.. All logs are already off and I'm just sending one query directly from the application (from Visual Foxpro using SQL Pass through and in VB.net using ado.net). I've found that the "probl

Re: [SQL] Extremely Low performance with ODBC

2008-05-23 Thread Richard Broersma
On Fri, May 23, 2008 at 9:47 AM, Sebastian Rychter <[EMAIL PROTECTED]> wrote: > Thanks.. All logs are already off and I'm just sending one query directly > from the application (from Visual Foxpro using SQL Pass through and in > VB.net using ado.net). I've found that the "problem" is that I'm using

Re: [SQL] Extremely Low performance with ODBC

2008-05-23 Thread Sebastian Rychter
:04 p.m. Para: Sebastian Rychter CC: pgsql-sql@postgresql.org Asunto: Re: [SQL] Extremely Low performance with ODBC On Thu, May 22, 2008 at 11:29 AM, Sebastian Rychter <[EMAIL PROTECTED]> wrote: > Hi, I'm executing a query through psql ODBC which is taking around 2 minutes > to compl

Re: [SQL] Extremely Low performance with ODBC

2008-05-23 Thread Richard Broersma
On Thu, May 22, 2008 at 11:29 AM, Sebastian Rychter <[EMAIL PROTECTED]> wrote: > Hi, I'm executing a query through psql ODBC which is taking around 2 minutes > to complete. When I run it from PgAdmin it takes less than 3 seconds. > > Any idea ? 1) turn off all ODBC query logging from both the ODBC

Re: [SQL] Extremely Low performance with ODBC

2008-05-23 Thread Craig Ringer
Sebastian Rychter wrote: Hi, I'm executing a query through psql ODBC which is taking around 2 minutes to complete. When I run it from PgAdmin it takes less than 3 seconds. The query itself has : . 15 inner joins (from just around 10 different tables - the other inner joins are using different a

Re: [SQL] Extremely Low performance with ODBC

2008-05-22 Thread Steve Crawford
Sebastian Rychter wrote: Hi, I'm executing a query through psql ODBC which is taking around 2 minutes to complete. When I run it from PgAdmin it takes less than 3 seconds. The query itself has : . 15 inner joins (from just around 10 different tables -- the other inner joins are using diffe