Re: Archivelog Mode Recoverability: How Close to Point of Failure?

"Ron" <support_at_dbainfopower.com> wrote in message news:g7adnVobveHArbHdRVn-hg_at_comcast.com...
> Hello Howard,
>
> I assume WCS is that all database data are lost :-)

Well, I don't want to drag this out, but that's not what the original poster was asking about (which I think is what our answers ought to be directed towards). And even then...

> Unless standby/replication architecture is in place - backup is what is
> left.

...you would of course have backups of your archive logs, and would thus be able to restore the database files and roll it forward some.

What I'm getting at is that it is incorrect to advise someone that the worst case scenario is to restore to the time of their last backup, because that is only true for someone who is running in noarchivelog mode.

What's more, if you lost your entire database (control files, online redo logs and data files) AND lost every single one of your archives since the time of your last backup, then it is also true to say that your last backup is unusable. Taken hot, it would be internally inconsistent, and would require the application of redo that is contained in the archives you've lost. The true worst-case scenario would therefore be that you'd have to restore from the backup *before last*. One can then dream up even "worser" worst case scenarios from there.

If this seems picky, then I apologise. But backup and recovery is a tricky subject and careless talk costs lives... well, data, anyway. Precision in the matter is therefore of prime importance.

Regards
HJR

-- 
--------------------------------------------
Oracle Insights: www.dizwell.com
--------------------------------------------