Requesting 2nd opinion on 10g backup strategy

Hi,

I usually dont post in more than one forum but since I didnt receive a response on the other forum i am posting it here too.

I have a 10G R2 database running on RHL 4.0 AS.

The flash recovery area is pointing to a windows server on our network having over a Terabyte of storage. This has been mounted using cifs.

I have the following RMAN script which will take a daily full database backup of our database to the windows shared disk (flash recovery area).

connect target sys/passwd_at_targetsid
connect catalog rman/passwd_at_catsid

CONFIGURE CONTROLFILE AUTOBACKUP ON;
CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 14 DAYS; run {
allocate channel ch1 type disk maxpiecesize=2G; backup database plus archivelog;
BACKUP ARCHIVELOG UNTIL TIME 'SYSDATE-7' delete input;

}
delete obsolete;

The strategy is as follows

1. The backups are taken to the disk and are picked up by tape on a daily basis.
2. At any given point of time there is only 1 backup (last) available on the disk. Older ones are deleted using an OS command on a dily basis.
3. The tapes are recycled every 14 days hence our retention policy is 14 days after which RMAN will delete the backups from its catalog if it needs to.

I would really appreciate critical/non critical feedback on this strategy. Am I missing something here.

One problem that we have though is that since we are using flash recovery area oracle is doing space management and occupies the space for 14 days backups even though they have been backedup on tape and deleted from the OS. I know that oracle have a backup of backupset option but thats not useful for us as I dont manage the tapes. I am looking for a way to have the Flash recovery area only occupy the space for 1 backup but still be able to have the 13 addiional backups record in the recovery and restore/recover when the need be.

Please help.

Thanks
Ron Received on Tue Apr 25 2006 - 13:48:06 CDT