Re: [martin@bugs.unl.edu.ar: BUG in logs]

I have looked at this behavior, which indeed is new for 8.1. The change
was caused by code I think I did to improve the behavior of
log_statement, specifically streamlining how we check for the type of
command.

In looking at reverting to the 8.0 behavior of logging error commands
with 'all', I see it is going to be hard to do, specifically since we
added behavior of logging the PREPARE query when EXECUTE is sent. I
don't think we want to lose that feature, and to have it we have to
first parse the statement, with possible exit on error.

What I have done is to document that errors are not output by
log_statement, and added as suggestion to use log_min_error_statement
for this purpose. I also fixed the code so the first EXECUTE has it's
prepare, rather than the last which is what was in the current code.

I also removed the "protocol" prefix from the PREPARE output, because in
fact both protocol and SQL-level prepares can be executed by SQL
EXECUTE.

Patch attached. I have backpatched this to 8.1.X.

---------------------------------------------------------------------------

Martin Marques wrote:
>
> OK, you're right about the log_min_error_statement value, but this behaviour has changed from 8.0. In earlier versions ERROR statements did get logged if log_statment was set to all or in 7.4, set to "on" DMaybe I missed something in the changelog of 8.1?
>
> On Tue, 11 Apr 2006 23:51:51 +0200, "Guillaume Smet" <guillaume(dot)smet(at)gmail(dot)com> wrote:
> >> From: Martin Marques <martin(at)bugs(dot)unl(dot)edu(dot)ar>
> >> I encountered a rare BUG in the way PG is logging. Let me first enlight
> > with some configuration I have and PG version:
> >
> > Perhaps I'm missing something but I think it's not a bug but a
> > configuration problem.
> >
> >> log_min_error_statement | panic
> >
> > If you set this one to error instead of panic, you will have your
> > failed statements logged.
> >
> >> log_statement | all
> >
> > This one only logs successful queries so it's normal you don't have
> > the statement in the log file if it fails.
> >
> > Regards,
> >
> >
> --
> ---------------------------------------------------------
> Lic. Mart?n Marqu?s | SELECT 'mmarques' ||
> Centro de Telem?tica | '@' || 'unl.edu.ar';
> Universidad Nacional | DBA, Programador,
> del Litoral | Administrador
> ---------------------------------------------------------
>
>
>
> ---------------------------(end of broadcast)---------------------------
> TIP 9: In versions below 8.0, the planner will ignore your desire to
> choose an index scan if your joining column's datatypes do not
> match
>

--
Bruce Momjian http://candle.pha.pa.us
EnterpriseDB http://www.enterprisedb.com

+ If your life is a hard drive, Christ can be your backup. +