Re: FW: idl_ub2$ and utlip.sql

From: Ujang Jaenudin <ujang.jaenudin_at_gmail.com>
Date: Thu, 26 Jul 2007 11:58:35 +0700
Message-ID: <3edcb66e0707252158v5a1a1fe6g70923bde28e0a3b4@mail.gmail.com>

mark,

db is 9.2.0.4 on linux x86
yesterday i was fixed it by

- changing init prameter,
- truncate idl_ub2$ drop storage
- execute utlirp twice
- then utlrp,
- to make sure..... then change parameter to original what oracle apps

required,
- and rerun utlrp, so far it is OK,

but got 69 packages from APPS invalidated.

checking the corruption, is OK, it seem physical corrupt. and now IDL_UB2$ has no corruption issue.

btw, does anyone have the experience the same and in Oracle Apps environment? if I export import Full database, is it any issues (due to oracle Apps environment).

regrds
ujang

On 7/25/07, Mark W. Farnham <mwf_at_rsiz.com> wrote:
> I'm curious about your storage allocations for the broken object(s).
>
> These are originally built based on sql.bsq but also depend on your block
> size and other choices about your storage.
>
> What is your Oracle release, block size, choices about local versus
> dictionary managed system and sysaux tablespaces, and file storage (file
> system, raw volumes, or ASM)?
>
> While unlikely, it is possible that you have a combination of space required
> from the construction of your total of PL/SQL that makes it impossible to
> store within the configuration constraints. If the bulk of your objects in
> SYSTEM still function correctly it certainly makes one wonder why idl_ub2$
> would be broken and whether it is the underlying storage or a logical
> corruption.
>
> Regards,
>
> mwf
>
> -----Original Message-----
> From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org]
> On Behalf Of Ujang Jaenudin
> Sent: Wednesday, July 25, 2007 5:13 AM
> To: Hallas, John (EXP N-ARM)
> Cc: oracle-l_at_freelists.org
> Subject: Re: idl_ub2$ and utlip.sql
>
> ya i tried also dbms_repair, check_object, fix and skip procedures
> but it seem realy physical corruption.
>
> oracle support told me to export import full database.
> is there any other shortcut way?
>
> regards
> ujang
>
> On 7/25/07, Hallas, John (EXP N-ARM) <john.hallas_at_lmco.com> wrote:
> > DESCRIPTION
> > Rem This script can be used to invalidate all existing PL/SQL
> > modules
> > Rem (procedures, functions, packages, types, triggers, views) in a
> > Rem database so that they will be forced to be recompiled later on
> > Rem either automatically or deliberately.
> > Rem
> > Rem This script must be run when it is necessary to regenerate the
> > Rem compiled code because an action was taken that caused the old
> > code's
> > Rem format to be inconsistent with what it's supposed to be, e.g.,
> > when
> > Rem migrating a 32 bit database to a 64 bit database or vice-versa.
> > Rem
> > Rem STEPS:
> > Rem
> > Rem (I) Invalidate all stored PL/SQL units (procedures, functions,
> > Rem packages, types, triggers).
> > Rem
> > Rem (II) Reload PL/SQL package STANDARD and package DBMS_STANDARD.
> > Rem
> > Rem (III) Selectively invalidate views and synonyms
> > Rem
> >
> > Doesn't seem high risk as long as you recompile everything afterwards.
> > Use utlrp for that
> > Why not log a SR
> >
> > You could try and fix the corrupt blocks with dbms_repair.check_object
> > procedure and dbms_repair.fix_corrupt_blocks
> >
> > More importantly, how has the corruption occurred
> >
> > John
> >
> > -----Original Message-----
> > From: oracle-l-bounce_at_freelists.org
> > [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Ujang Jaenudin
> > Sent: 25 July 2007 04:15
> > To: oracle-l_at_freelists.org
> > Subject: idl_ub2$ and utlip.sql
> >
> > folks,
> >
> > I have problem on system tablespace, corruption on object idl_ub2$,
> > searching google that idl_ub2$ can rebuild by utlip.sql, is it true?
> >
> >
>
>
> --
> regards
> ujang
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>
>
>

-- 
regards
ujang
--
http://www.freelists.org/webpage/oracle-l

Received on Wed Jul 25 2007 - 23:58:35 CDT