Re: Data Guard question.
- From: Alex Gorbachev <gorbyx@xxxxxxxxx>
- To: RROGERS@xxxxxxxxxxxxx
- Date: Sun, 29 May 2005 12:50:14 +0200
See inline.
> 1. Using a physical database as the Data Guard, does the second server =
=3D
> have to have the exact same physical equipment for the disks or does it =
=3D
> just have to be the same directory structure? The reason I am asking is =
=3D
> the production server has a SAN that I use as /U03 mount point where the =
=3D
> archivelogs and a controlfile are kept. I can not get another SAN for the=
=3D
> standby server but would like to use a SCSI box with the same mount point=
=3D
> but different disk size.
As people mentioned it's not required. However, having different
hardware on standby site you should accept that performance of you
database might be worse after failover and that you might be in some
kind of degraded mode after disaster. This should match to your SLA.
> 2. Currently I use RMAN nocatalog mode for the backups of the production =
=3D
> database to disk. It appears from what I have read the when I use RMAN an=
d =3D
> Data Guard I have to use a catalog. IS this a true fact? Can you point to=
=3D
> a doc that clears this question up?
You don't have to use RMAN at all. RMAN with catalog is convenient for
GAP resolution if you want to restore archived logs to standby but you
need to access RMAN backup files anyway. Another way is to restore it
on primary and let them to be delivered on FAL requests or copy them
manually on standby and register there (ALTER DATABASE REGISTER
PHYSICAL LOGFILE).
--=20
Best regards,
Alex Gorbachev
--
//www.freelists.org/webpage/oracle-l
Other related posts:
