Re: Noarchivelog 11.2.0.1 Restart
Date: Wed, 26 Jan 2011 15:17:17 -0800 (PST)
Message-ID: <738263.12359.qm_at_web80203.mail.mud.yahoo.com>
After a bunch more thinking and reading I found something that works. shutdown immediate;
startup nomount;
restore controlfile from 'C:\oracle11g\flash_recovery_area\orcl\backupset\2010_12_21\O1_MF_NCSNF_TAG20101221T175022_6k2D4NSM_.BKP'; startup mount;
startup open resetlogs;
In short, it appears that I had to specify the location of the backup piece that contained the controlfile backup even though list backup could see it. It appears that backup database under most scenarios will backup the control files. Autobackup did not help.
So now, I need to write a script such that all future restores are point and shoot as much as I've enjoyed this experience.
Thank you all for the great assistance and kind indulgence. Edward Long
- On Wed, 1/26/11, Ed Long <rdhm99a_at_prodigy.net> wrote:
From: Ed Long <rdhm99a_at_prodigy.net>
Subject: Re: Noarchivelog 11.2.0.1 Restart
To: "De DBA" <dedba_at_tpg.com.au>
Cc: "'oracle-l_at_freelists.org'" <oracle-l_at_freelists.org>
Date: Wednesday, January 26, 2011, 1:30 PM
Hi everyone.
Here is today's update.
Oracle seconded "De DBA"''s suggestion to recreate the controlfile referring to document 735106.1 for specifics.
Step 1 is alter database backup controlfile to trace.
Step 1 fails with ORA-16433 database m/b in r/w mode.
I now believe this bug is Windows 7 related as that is what is different.
Any experience with 11.2.0.2?
Edward Long
- On Wed, 1/26/11, De DBA <dedba_at_tpg.com.au> wrote:
From: De DBA <dedba_at_tpg.com.au>
Subject: Re: Noarchivelog 11.2.0.1 Restart
To: "Ed Long" <rdhm99a_at_prodigy.net>
Cc: "'oracle-l_at_freelists.org'" <oracle-l_at_freelists.org>
Date: Wednesday, January 26, 2011, 5:01 AM
That could be your problem then: RMAN's "restore database" command does not restore the control file or the SP-file. It's got to be done explicitly.
The existing control file after the RAT "knows" that the database has had activity after your level 0 backup and RMAN wants to bring the restored database up to speed. Hence the message that file 1 needs media recovery.
Cheers,
Tony
Ed Long wrote:
Thanks for thinking about the problem.
No, I literally just did restore database.
Ticket open with Oracle, more updates tomorrow (Tuesday).
Edward Long
- On Mon, 1/24/11, De DBA <dedba_at_tpg.com.au> wrote:
From: De DBA <dedba_at_tpg.com.au>
Subject: Re: Noarchivelog 11.2.0.1 Restart
To: "'rdhm99a_at_prodigy.net'" <rdhm99a_at_prodigy.net>
Cc: "'oracle-l_at_freelists.org'" <oracle-l_at_freelists.org>
Date: Monday, January 24, 2011, 7:57 PM
Hi Ed,
Perhaps a silly question, but did you restore the control files after completing the RAT? If not, the post-RAT control file will have a higher SCN than the restored database and insist that the datafiles need media recovery. There should be an automatic controlfile backup in your level 0 backup, since it contains the system datafile.
Cheers,
Tony
-- http://www.freelists.org/webpage/oracle-lReceived on Wed Jan 26 2011 - 17:17:17 CST