Skip to Main Content

Database Software

Announcement

For appeals, questions and feedback about Oracle Forums, please email oracle-forums-moderators_us@oracle.com. Technical questions should be asked in the appropriate category. Thank you!

RMAN duplicate problem ORA-01547, different sid, different box

FunkyJan 10 2009 — edited Jan 14 2009

Hi!

we have used for a long time DUPLICATE for daily backup/restore checking our PROD instance on other instance (different sid - different box). Our RMAN backup is on daily basis and full (no increment).

Backup starts every day at 02:00 and finishes about 04:10. Naturally we use catalog database (CATDB) on different server.

For archive system we use Legato 7.2 version. All databases (10.2.0.2) are on Linux RH EE 4.5 64 bit. Catalog db (CATDB) is on XE 2.0 (free version) on Linux RH EE 4.5 32 bit.

Statistics are based on weekly level on CATDB and PROD (target) according all books ...

All looks by the book....

For a long time we had no bigger problems (sometimes duplicate failed but we thought no need to worry for "some" problems). But recently we got more unsuccessful duplicates in a short period.

In file is a log of such unsuccessful duplicate:
http://rapidshare.com/files/181623335/duplicate_problem.txt.html

All the time it looks like when recover part should be started, there are no archived log files in backup!?

If I try immediately after to restore only archive logs:

RUN {
  allocate channel t1 type 'SBT_TAPE';
  send 'NSR_ENV=(NSR_SERVER=zzz.hypohr,NSR_CLIENT=xxy)';
  SET ARCHIVELOG DESTINATION TO '/u01/lun0/dolly/arch';
  RESTORE ARCHIVELOG TIME BETWEEN "TO_DATE('2008-01-09 02:00:00','yyyy-mm-dd hh24:mi:ss')" AND "TO_DATE('2008-01-09 04:40:00','yyyy-mm-dd hh24:mi:ss')";
  release channel t1;
}

I get all needed archive logs to manually recover database (recover database using backup controlfile until...).

I was validating backup and looks perfect....(may sent the log).

I was researching the problem and found that really exists some notified errors in our catalog database (alert log of target instance as well from auxiliary looks OK).

In RMAN.RSR table I get records that were suffering problems. Attached log is based on following sql:

SELECT 
   ROWID, R.RSR_KEY, R.DBINC_KEY, R.RSR_RECID, 
   R.RSR_STAMP, R.RSR_PKEY, R.RSR_L0KEY, 
   R.RSR_LEVEL, R.RSR_TYPE, R.RSR_OPER, 
   R.RSR_CMDID, R.RSR_STATUS, R.RSR_MBYTES, 
   R.RSR_START, R.RSR_END, R.RSR_IBYTES, 
   R.RSR_OBYTES, R.RSR_OPTIMIZED, R.RSR_OTYPE, 
   R.RSR_SRECID, R.RSR_SSTAMP, R.RSR_ODEVTYPE
FROM RMAN.RSR R
Where
RSR_OPER = 'RMAN'
 AND RSR_STATUS like '%ERROR%'

http://rapidshare.com/files/181625271/RMAN_rsr_errors.txt.html

I have tried to duplicate by the hand (restore datafiles through RMAN, restore archived logs from RMAN, recover through sqlplus) all works like like a charm!

Is there any way that someone point me where to start digging to find out why some duplicates terminated with errors.

THX

This post has been answered by ebrian on Jan 14 2009
Jump to Answer
Comments
Locked Post
New comments cannot be posted to this locked post.
Post Details
Locked on Feb 11 2009
Added on Jan 10 2009
15 comments
3,025 views