ORA-1861 possible controlfile corruption after RMAN restore

From: Rich Jesse <rjoralist2_at_society.servebeer.com>
Date: Fri, 29 Jun 2012 11:13:06 -0500 (CDT)
Message-ID: <75d19ddcb7083673eb5a384e106e273b.squirrel_at_society.servebeer.com>

Hey all,

I've copied RMAN backup files from our production 10.1.0.5 DB (AIX 5.3 TL8) to our dev box (AIX 5.3 TL12). I've been restoring and destroying this particular DB on a dev box dozens of times using the same RMAN files I copied almost a month ago. I start the restore with these commands (note that we do not use an RMAN catalog):

$ rman target /

RMAN> startup nomount;
RMAN> restore controlfile from '/u99/bleah/c-x909yaddayadda';
RMAN> alter database mount;
RMAN> catalog backuppiece '/u99/bleah/3mnc79ba_1_1';

(repeat for all other RMAN backup pieces) RMAN> delete expired backup;

However, yesterday on my second restore of the day, I received this after the delete completed:

RMAN-08132: WARNING: cannot refresh aged file list
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03009: failure of REFAF command on default channel at 06/29/2012 09:56:11

ORA-00604: error occurred at recursive SQL level 1 ORA-01861: literal does not match format string

After several iterations of shutting down, wiping the test DB, and recreating, I'm still at a loss to explain what happened. There's very little reference to the REFAF error, although I suspect that's a symptom of the ORA-1861 instead of the issue.

Thinking my Windows copy of the RMAN commands that I copy/pasted somehow got a funky UTF8 quote in there, I manually typed the commands, to the same error. I also attempted to "catalog start with '/u88/bleah';", with no change.

I dumped the controlfile using the session event 'immediate trace name controlf level 20', but I don't know exactly what to look for in the resulting trace. I compared the traces from before and after the catalog commands, but again, I'm not sure what changes are important and what's not.

SOMETHING changed yesterday, but I'll be durned if I can figure out what it is. At this point, I'm considering physical or logical corruption of the RMAN controlfile backup on the JFS2 filesystem -- what else could it be?

Thoughts anyone?

TIA!
Rich

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

Received on Fri Jun 29 2012 - 11:13:06 CDT