Re: Rman logfile restore and catalog behaviour

On Jul 5, 2:52 pm, Neil Cudd <neil.c..._at_blueyonder.co.uk> wrote:
> Hi there,
>
> this could get confusing so bare with me :
> AIX 5.3 Oracle 9.2.0.7
>
> It's been decided to move the datacentre half way around the world
> with minimal downtime.
> We are backing up using RMAN to a datacentre in this hemishpere.
>
> So the backup tapes were exported and sent to far away to be imported
> to the tape server on the remote site (too big to be sent down the
> wan).
>
> The database is restored at the far side and put in recovery mode,
> rolling forward, catching up on the logfiles until the cutover date
> when the local DB server is shutdown - 30 days time.
>
> So I am restoring archivelogs - via rman - from 5 days ago to send
> down the wire to be applied to remote copy.
>
> However, our daily archive log rman job is now complaining that it
> cannot find log sequence #### - this being the first of the old
> archive logs that I restored localy to send down the wire. (I restored
> then deleted - not wanting to crowd the archive log area.
>
> What am I doing wrong ? (no asides on startegy - I didn't decide it,
> just left to apply it).
>
> Many thanks for any help.
> Neil

It's all here
I restored
> then deleted - not wanting to crowd the archive log area.

One basic rule with RMAN is you should never ever delete a log file using the O/S. You should leave this to RMAN. You could set up a recovery window, after which you could use RMAN to run the DELETE OBSOLETE command, which also deletes archivelogs (and deletes them from the O/S),
or you set up a scheme to backup your archives multiple times and use the delete input qualifier, which would also delete them. You can do this in various ways, but do NOT use the OS to delete archivelogs.

--
Sybrand Bakker
Senior Oracle DBA