Re: oracle 8i and redo logs during recovery.

Hopefully your BCVs also contain the current controlfile. You can just put all the archived redologs, online redologs and current controlfiles in place on the backup server copy of the database, and recover completely. When the recover database command runs out of archived redologs, it will switch into the onlines as though it is performing crash recovery.

I'm not sure why with all this hardware you are not running a standby. It seems like a lot of trouble to automate your daily copy, when you could have an up-to-date copy always recovering just minutes behind the primary. With your current system your MTR takes as much time as is required to apply all logs for up to a 24-hour period.

Anyway, best of luck,
--
Jeremiah

Steve Perry wrote:
>
> 2 servers running 8.1.5 on Solaris 2.6, EMC disk arrays
> At night a hot backup is done on host 'a' and then copied to host 'b'
> (backup server)
> middle of next day, host 'a' crashes
> We have all the archive logs and redo logs (established BCV's)
> We connect the BCV from above to host 'b' and start the database
> recover database
>
> Question: When oracle applies the archive logs, what happens to the redo
> logs, specifically the redo log that was active at time of the disaster?
> I didn't see any information in the documentation that explains this process
> in detail.
Received on Tue Dec 28 1999 - 12:02:09 CST