Re: Stanby cannot be moved forward

From: <fitzjarrell_at_cox.net>
Date: 30 Oct 2006 11:55:19 -0800
Message-ID: <1162238119.048548.8760@k70g2000cwa.googlegroups.com>

Comments embedded.
Massa Batheli wrote:
> Let me rephrase and could be lucky get a response.
>
> There are three servers.
> 1)Production
> 2)Standby
> 3)Backup

Okay.

> A backup is taken from production server to backup server.
> Recovered and opened read only to acertain the validity of backup.

Then this isn't a proper backup for standby instantiation.

> This is then moved to standby server and ready for logs to be applied
> Is there any problem with this?

Many. You need to be doing this through rman to create a backup suitable to create a standby database:

 RUN {
 # Backup Database
 delete noprompt obsolete redundancy 1 DEVICE TYPE DISK;  ALLOCATE CHANNEL ch00 TYPE DISK MAXPIECESIZE = 3G;  BACKUP

       FULL
       FORMAT '<backupdir>/bk_%d_%s_%p.bak'
       DATABASE
       include current controlfile for standby;

 sql 'alter system archive log current';  RELEASE CHANNEL ch00;
 }

Then move these pieces to the standby server, ensure connectivity between production and standby, connect to the production and catalog databases with rman:

rman target user/pass@<production db> catalog user/pass@<primary catalog db> auxiliary / msglog mylog append

then run:

run {
  set command id to 'Create Standby Database';   allocate auxiliary channel d1 type disk;   duplicate target database for standby

     NOFILENAMECHECK
     dorecover;

}

Of course this presumes you have the same filesystem configuration on both servers.

> The standby fails consistently

For obvious reason of there not being a standby controlfile.

>and there have been suggestions that you
> do not have to recover.
>
>
>

It isn't a standby database, it won't recover by that mechanism.

David Fitzjarrell Received on Mon Oct 30 2006 - 13:55:19 CST