Re: ORA-1861 possible controlfile corruption after RMAN restore
Date: Fri, 29 Jun 2012 12:06:30 -0500 (CDT)
Message-ID: <5400644343affb47472a403bba90937f.squirrel_at_society.servebeer.com>
Hey Alex,
> There were quite a bit of bugs with RMAN writing an incorrect date
> into the contorlfile (like Feb 30th). If that's your case then see if
> you have something like that in your trace.
I've read MOS 852723.1, and I didn't think it was a date issue since this query bombed consistently on the same deleted backup piece:
SQL> select * from v$backup_piece;
However, I was able to SELECT RECID FROM V$BACKUP_PIECE, as well as every column logically up to COMPLETION_TIME. I was also able to SELECT rows using RECID immediately before and after the last row *displayed* in the results of the "SELECT *". I narrowed down the RECIDs until I got to the one that generated the error -- the RECID of the row created by the CATALOG BACKUPPIECE. This tells me that it's the COMPLETION_TIME of the new backup piece. Another quick session trace of the controlfile and (now that I know what to look for) the resulting trace does indeed show this:
Latest record:
RECID #208910 Recno 1065 Record timestamp 06/29/12 11:45:49 piece #1 copy
#1 compressed #yes pool 0
Backup set key: stamp=784574397, count=209028
[blah blah blah]
Completion time 06/31/12 21:49:37
June 31. Rats.
Well, I know what's happening now. I just need to fix it...
Thanks Alex!
Rich
-- http://www.freelists.org/webpage/oracle-lReceived on Fri Jun 29 2012 - 12:06:30 CDT