Charles J. Fisher wrote:
> On Mon, 28 Feb 2005, Sybrand Bakker wrote:
>
>
>>*Which* infrastructure? You mean *NO* Infrastructure?
>
>
> Under 8i, you need a separate database server to host the recovery
> catalog, which means a separate license AFAIK. Not worth my trouble nor
> the expense.
>
>
>>I would be suspicious of any DBA who still, after more than 4 years of
>>RMAN being operational, rejects RMAN as a viable backup solution, as
>>you do.
>
>
> I run more Oracle 7 than 8. I have ABSOLUTELY NO INTENTION of relying upon
> Oracle's provided solution for v7 - the "Enterprise Backup Utility."
>
> I'm pretty suspicious of you... you must be a newbie to Oracle. :)
>
> For the few 8i instances that I maintain, RMAN simply isn't worth the
> trouble; a consistent backup approach is more rational.
>
>
>>I have seen third party backup sw bringing *ALL* tablespaces in hot
>>backup. Oracle advises against this for good reasons. I assume you
>>rather want to continue with this type of software?
>>Your company is really lucky when they discover you can't restore your
>>backup.
>
>
> Actually, that is a good approach if you are using a SAN for backups. If I
> were to have my StorageWorks systems configured to "RUN CLONE" then
> throwing the entire database into backup and using the firmware to copy is
> viable (and fast, which might result in less overall redo generation).
> Getting the right LVM configuration would be very tricky, though.
>
> I don't use any 3rd-party backup software (at long last - so long BMC).
>
> RMAN is probably a slightly more appropriate tool in later versions of
> Oracle. RMAN is nice in that it will skip free blocks, and it brings back
> the cumulative backups that were stripped from exp/imp.
>
> Every approach and tool has its uses, but to argue that one tool dominates
> all others is to do a disservice to every tool that you own.
>
> --------------------------------------------------------------------------
> / Charles J. Fisher | A bad design with a good presentation is doomed /
> / cfisher_at_rhadmin.org | eventually. A good design with a bad /
> / http://rhadmin.org | presentation is doomed immediately. /
> --------------------------------------------------------------------------
Sybrand is no newbie and suffers less than gracefully with some of the
paleolithic databases he is forced to support. But I can not imagine
anyone suggesting you upgrade from paleolithic (v7) to neolithic (v8i)
when neither is supported. Any upgrade worth doing today is to v10R1.
--
Daniel A. Morgan
University of Washington
damorgan_at_x.washington.edu
(replace 'x' with 'u' to respond)
Received on Tue Mar 01 2005 - 11:39:03 CST
