Re: changing from SS2005 to 11g?

I don't think the choice of Oracle release is especially relevant here. The real question is whether to migrate from one database engine to another.

The very first question to ask here, I think, is: "What do you propose to achieve by this?"

You are going to want some pretty solid reasons to undertake something like this. Assuming your application is non-trivial and has been coded to use SQLserver to good affect (i.e., that it is not "database agnostic"), you are going to have to recode -- possibly even redesign -- most of the application. You are going to need to show a very substantial benefit to justify this effort.

If the application is "trivial" -- i.e., it stresses neither the hardware nor the database engine -- there can be no benefit to migrating. At best, you will migrate from an idle SQLserver system to an idle Oracle system. At worst, you'll get something wrong, and things will go poorly on the new database.

If the application is "database agnostic", it probably already sucks. But it will probably such just as much on Oracle as it does on SQLserver. In this situation, a recoding/redesign effort could well be warranted, however, changing database platforms as part of this migration is likely to offer only one concrete benefit: it will force you to throw away more of the crappy old code.

Don't get me wrong on this. As an Oracle professional, I am always delighted to see people migrate to Oracle, from *anything*, really. It means more work for me and people like me. But the last thing I would want to see is people proclaiming that "Oracle sucks" on the basis of a poorly conceived or execute migration project. That's not good for anybody, except maybe Microsoft.

Just my $0.02 worth...

Oh, by the way -- having said all that, I can now say that there is probably nothing wrong with going with 11g for a project like this. If you do it right, you're probably going to spend a year or two on this migration by which time 11g should be more than ready for "prime time", and 10g will be about to be de-supported. If you don't plan to spend at least a year on this project, you may well be doomed to failure, so the relative "immaturity" of 11g could be a convenient excuse. It looks like 11g is an advantage either way. :-)

On Fri, Jul 4, 2008 at 3:08 PM, Andrew Kerber <andrew.kerber_at_gmail.com> wrote:

> Its too early to go to 11g. Let someone else do Oracle beta testing for
> them.
>
>
> On Fri, Jul 4, 2008 at 12:48 PM, Bjørn D. Jensen <bjorn.d.jensen_at_gmail.com>
> wrote:
>
>> Hi!
>> If your boss want to know if it is a good idea to migrate...
>>
>> I think it must be clear what the business purpose should be
>> and staff must be educated...
>>
>> What are your opinions/experiences?
>>
>> Best regards
>> OCP & MCTS Bjørn D. Jensen
>> --
>> http://www.freelists.org/webpage/oracle-l
>>
>>
>>
>
>
> --
> Andrew W. Kerber
>
> 'If at first you dont succeed, dont take up skydiving.'

-- 
Cheers,
-- Mark Brinsmead
Senior DBA,
The Pythian Group
http://www.pythian.com/blogs

--
http://www.freelists.org/webpage/oracle-l