Re: Oracle configuration related weird behaviour?

From: <qualitat8_at_gmail.com>
Date: Thu, 4 Jun 2009 02:16:49 -0700 (PDT)
Message-ID: <718a4f15-4e7f-476b-a778-6e121ab7218a_at_r34g2000vba.googlegroups.com>



I have no acces to Metalink for patches.

I'm using this version because we are doing maintenance tasks in an application running in 9.2.0.8, but we use the freely available version 9.2.0.1.

Is there any other information source without purchaising Oracle support where I can see if it's a declared bug?

Thank you.

sybrandb ha escrito:
> On 4 jun, 09:55, qualit..._at_gmail.com wrote:
> > I want to replicate some schemas from a Oracle 9.2.0.8 to a 9.2.0.1.
> > The export/import process was OK and now I've objects replicated and
> > data loaded.
> >
> > A function is not working properly, and an example of what's the
> > problem can be reproduced with this example:
> >
> > declare
> >     cursor cPrueba is
> >     select 'apruebas' as nom from dual
> >     union
> >     select trim('This is a very very long text' || ' ' || 'This is
> > another very very long text' ) as nom from dual
> >     order by nom;
> >     begin
> >     for elemento in cPrueba
> >     loop
> >     dbms_output.put_line('this is ' || elemento.nom);
> >     end loop;
> >     end;
> >
> > I've been trying with this query, and when I added the first select,
> > it generates a ORA-06502: PL/SQL: numeric or value error string.
> >
> > If I do a substr in the second select to truncate that text to be as
> > long as the first one  'apruebas' plus one,  8+1, then it works ok.
> > The rule applies if I change the first string, just substr the next
> > ones to be the first length + 1.
> >
> > It seems the first row determines the length of that 'column'.
> >
> > I haven't this problem in the first database, so this is the reason
> > why I think the diference may be related to server configuration.
> >
> > Is there any configuration parameter determining this behaviour?
> >
> > Thank you in advance.
>
> The problem is related to comparing a patched, desupported, release
> 9.2.0.8 with a non-patched, equally desupported, base release 9.2.0.1.
>
> This is exactly the type of issue being left-over in a base release
> and being patched in a patch release.
> So if it works, in 9.2.0.8, dump 9.2.0.1 or better still : dump 9.2
> and upgrade to 10gR2.
>
> --------
> Sybrand Bakker
> Senior Oracle DBA
>
> Experts : those who *did* read the documentation.
Received on Thu Jun 04 2009 - 04:16:49 CDT

Original text of this message