Re: sql_text during parsing
Date: Mon, 12 Jul 2021 16:40:35 +0100
Message-ID: <CACj1VR59pty-72xBCx44-5Xw-h-8WsvHjC6jFX-x5adQk0s5Kg_at_mail.gmail.com>
I could probably guess 99% of the statement ;)
Joking aside, if the full SQL didn’t dump out into the process’s trace file, you might be able to do some detective work based on the library locks the session has taken out.
Joining v$libcache_locks to v$db_object_cache on object_handle = addr where holding_user_session is your parser’s saddr. If you filter on namespace = 'TABLE/PROCEDURE', you should be able to see which tables the SQL needs to be looking at.
Hope that helps,
Andrew
On Mon, 12 Jul 2021 at 16:27, Stefan Koehler <contact_at_soocs.de> wrote:
> Hello Nenad,
> any chance with a process state dump (or error stack trace)?
>
> Best Regards
> Stefan Koehler
>
> Independent Oracle performance consultant and researcher
> Website: http://www.soocs.de
> Twitter: _at_OracleSK<
>
> > Noveljic Nenad <nenad.noveljic_at_vontobel.com> hat am 12.07.2021 17:25
> geschrieben:
> >
> > Thanks!
> >
> > I got a part of SQL (it’s chopped)
> >
> > Any idea for getting the full text?
> >
> > SQL> oradebug setospid 17876
> > Oracle pid: 107, Unix process pid: 17876, image: oracle_at_svdbp02p
> > SQL> oradebug current_sql
> > SELECT (CASE WHEN ((CASE WHEN (CASE WHEN (CASE WHEN (CASE WHEN (CASE
> WHEN (CASE WHEN (CASE WHEN (CASE WHEN ((CASE WHEN ((CASE WHEN (((CASE WHEN …
> >
> > The problem might be with the parsing of a nested case.
> >
> > Have to get the full text to experiment.
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>
-- http://www.freelists.org/webpage/oracle-lReceived on Mon Jul 12 2021 - 17:40:35 CEST