RE: hot-backup recovery question

> -----Original Message-----
> From: Hollis, Les [mailto:Les.Hollis_at_ps.net]
> Sent: Monday, March 14, 2005 10:40 AM
> To: holland_at_guidancetech.com; hkchital_at_singnet.com.sg; oracle-l_at_freelists.org
> Subject: RE: hot-backup recovery question
>
> We do this very thing every night plus we rename the database and add
> indexes for reporting purposes.
>
> Are you copying over any logfiles and applying them before you try to
> open? I didn't see that in your steps.
>
> You will need to copy over the logfiles n use when yopu started the
> alter tablespace begin backup, any logfiles generated during the mirror
> copy and the last logfile generated . When you do the alter tablespace
> end backup, immediately do an alter system switch logfile and copy that
> one over as well.
>
> Apply the logfiles and after the last one is applied then type cancel;
>
> Then you can do your alter database open resetlogs.

[...]
6. Verify they are all out of backup mode via:

        select count(*) from v$backup where status = 'ACTIVE'; 7. Archive the current redo log via:

        alter system archive log current;
8. Split the archived redo log file system's mirrors.

Step #7 switches log files AFTER the database has been taken out of backup mode, and then the archived log file system is split in Step #8 -- so I have the archived redo log with the "end backup" statements in it, along with all the other log files from before that point in time....

On the target system, the log files are replayed via:

        alter database recover automatic using backup controlfile until cancel.

When I reach the archived redo log which has the 'End backup' in it, I'm prompted for the next log file (which doesn't exist -- it hasn't been archived on the production system yet!), and when I tell it to 'cancel' recovery, I get the error about OPEN RESETLOGS needing more media recovery to be consistent...

I've seen this in the past on 9i databases if you get the "recover" syntax out of order (i.e. "recover until cancel using backup controlfile" will cause it, where "recover using backup controlfile until cancel" wouldn't).

Can you send me the syntax of your recovery command? Maybe something's changed between 8i and 9i -- I'm using the same syntax I successfully used to clone 9i systems (with a SID change).

Thanks!
Rich

--
Rich Holland        (913) 645-1950        SAP Technical Consultant
print unpack("u","92G5S\=\"!A;F]T:&5R(\'!E<FP\@:&%C:V5R\"\@\`\`");


--
http://www.freelists.org/webpage/oracle-l