Hi,

On Monday we started receiving this error in our backup log:

RMAN-00571:
================================================== =========
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS
===============
RMAN-00571:
================================================== =========
RMAN-03009: failure of delete command on d4 channel at 01/20/2012
02:31:24
ORA-19633: control file record 6197 is out of sync with recovery
catalog


We do not use a recovery catalog, we use the control file. Oracle
has been less than helpful, telling us that the only way to solve this
bug is to use a recovery catalog. I find that funny as we have 20
databases, each using their control file, and this is the first time
it has ever happened.

At any rate, I cannot find a solution. Can anyone help? I'm also
thinking if I can just delete all the backups, will it clear those
records from the control file, and I can start with a fresh level 0
backup.

Thanks!

"ExecMan" <artmerar (AT) yahoo (DOT) com> a écrit dans le message de news: 62c5b8b6-2c49-4ed7-9e8d-9d5603dada46...oglegroups.com...
ORA-19633: control file record %s is out of sync with recovery catalog
*Cause: The control file record describing the file to be deleted in
a call to deleteBackupPiece, deleteDataFilecopy, proxyDelete or
deleteArchivedLog does not match the validation data supplied by
recovery manager.
*Action: contact Oracle support


Regards
Michel

On Jan 20, 1:53*pm, "Michel Cadot" <micadot{at}altern{dot}org> wrote:
Gee, thanks. But I already said Oracle is of no help. Their
response to me: "Use a recovery catalog". Big help there.

I'm just wondering if anyone else has had this issue and found a
solution. Or, will deleting all of my RMAN backups and doing a fresh
level 0 fix it?

On Fri, 20 Jan 2012 13:16:54 -0800, ExecMan wrote:

I am surprised that Oracle would say so, there must have been a
misunderstanding. You can also try synchronizing things by using crosscheck
backup of database.



--
http://mgogala.freehostia.com

On Jan 20, 11:38*am, ExecMan <artme... (AT) yahoo (DOT) com> wrote:
Try doing a cross check of everything ... archive logs backups
etc ...
#!/bin/ksh
.. /home/oracle/ora_11g_env
export ORACLE_SID=xyz123 /* fix this sid */
rman << EOF
connect target
run {
allocate channel d1 type disk;
crosscheck backup;
crosscheck copy;
crosscheck backup of database;
crosscheck backup of controlfile;
crosscheck archivelog all;
release channel d1;
}
exit;
EOF

Something like the above should be reasonably close ... make sure it
executes cleanly ...

On Jan 20, 8:38*am, ExecMan <artme... (AT) yahoo (DOT) com> wrote:
See MOS Bug 6402384: ORA-19633: CONTROL FILE RECORD 765 IS OUT OF SYNC
WITH RECOVERY CATALOG
"Delete and recreate the controlfile which the customer does not want
to do"
It appears the controlfile records for backup pieces are reused before
those of backup sets, allowing duplicate set names.
I notice this is fixed in 10.2.0.5 and 11.1.0.7 patch sets.
What a choice: create a catalog, which would let you fix the
controlfile records through syncronization or re-create the
controlfile, with incarnation issues. Anything else would be scarier.

jg
--
@home.com is bogus.
http://www.extremetech.com/computing...d-service-dies

On Jan 20, 7:58*pm, joel garry <joel-ga... (AT) home (DOT) com> wrote:
When I searched Oracle support for the bug number I found 3 patches
for this issue. I did not look into the patch details, but a
potential fix may be available depending on which specific versions/
platforms the patches were for.

HTH -- Mark D Powell --

On Jan 20, 6:14*pm, John Hurley <hurleyjo... (AT) yahoo (DOT) com> wrote:
To add to this, you might attempt to move your backups and archivelogs
out of the FRA area (db_recovery_file_dest) and run a crosscheck +
delete expired and then move the files back to their original location
and re-catalog them. (catalog start with '/../.../' )

recreating the controlfile may be your only option. I have had to do
that recently due to a corrupted controlfile that was purely self-
inflicted more than a 6 months before I joined the team.