Re: Migrating from 8i

From: joel garry <joel-garry_at_home.com>
Date: Wed, 26 Aug 2009 14:39:58 -0700 (PDT)
Message-ID: <8f934c2e-998a-46cd-9412-cc64918c6581_at_r24g2000prf.googlegroups.com>



On Aug 26, 11:02 am, "Fabrice" <emouc..._at_spaminfonietest.fr> wrote:
> Hello,
>
> We are using Oracle 8i on the occasion of our professional application.
> Today we have 2 solutions :
>
> - migrate to an another application running with Sql Server.
>
> or
>
> - consolidate (for few years) our actual solution by migrating Oracle 8i to
> a new version : It seems to be cheaper for us.
>
> But I'm wondering to which version migrate : 9i, 10g or 11g.
>
> Is it possible to migrate a complete schema from 8i to any of theses
> versions : data, functions, triggers ....
> I have heard that the standard version is free. Is it true and if yes what
> are the limitations.
>
> Thnaks for your help.
> Fabrice.

I agree with John. I'd add, there are different upgrade paths (see the upgrade guide), if possible you want to export/import rather than upgrade in place. How appropriate that last sentence is depends on how you got to the current 8i, what you want to avoid are things like maintaining old DMT tablespaces.

I'd also add, you may have a lot of application code that is based on older ways of doing things, or even myths, that may or may not impact you, and may or may not show up in testing. Perhaps the biggest of these is if rule-based-optimizer assumptions are made in the code.

See http://www.oracle.com/technology/products/bi/db/10g/pdf/twp_bidw_optimizer_10gr2_0208.pdf http://www.oracle.com/technology/products/database/oracle11g/upgrade/presentations/9i_to_11g_real_world_customer_experience.pdf

My own experience (upgrading the same third party ERP/MRP apps) has been 8 to 8i was worth it, 8i to 9i was worth it, and 9i to 10g was worth it, even with some nasty forced upgrades in hardware and OS in some parts. Haven't done 11 yet. I'm still running into sudden optimizer craziness at odd rare times, but with the tools and information available, they've been few and easily solved (with some minor exceptions that haven't been show-stoppers). Every time I look at SQL-server I just shake my head, it looks like myths-and-legend from ye olde tymes, people rebuilding indices for no good reason... or worse, for good reason...

jg

--
_at_home.com is bogus.
http://revealingerrors.com/
Received on Wed Aug 26 2009 - 16:39:58 CDT

Original text of this message