Re: Strategy question for 24X7, how to do cold backups?

On Wed, 26 Sep 2001, richard.armstrong_at_juno.demon.co.uk wrote:

> You don't need to keep all of the archived redo logs from day
> one. Database changes are flushed from memory to disk
> regularly and you only need to keep archived logs containing
> changes that that have not yet been applied to the data files
> on your latest backup.

Well, if one is 24x7, how does anything other than the first cold backup from day #1 actually exist?

> Standby database is very useful but I believe the standby
> instance can't normally be used so the standby machine is idle
> during normal operation.

In this scenario, I would think that a standby database could be used to actually do the cold backups after day #1 as well as applying redo logs forward for so that a downed instance can come back online very quickly by switching to the standby. Which, I would assume, at that time becomes the primary and once the downed instance is up, this becomes the secondary.

> Have you considered other options such as replication or
> parallel server to maintain a duplicate database running on a
> separate server? These are more complicated than standby
> database but they allow all of your server resources to be used
> at the same time and your database is still available if one of
> the servers fails.

No, I'm just curious at how to understand this panacea of 24x7.

-- 
Galen Boyer
Everyone in town, now, they probably all agree,
I'm _lying_ in the bed I made.