Re: Deadly sins againts database performance/scalability

From: Galen Boyer <galenboyer_at_hotpop.com>
Date: 16 Dec 2003 09:36:09 -0600
Message-ID: <ur7z4g4lh.fsf@standardandpoors.com>

On Tue, 16 Dec 2003, paul_at_not.a.chance.ie wrote:
>
>
> Mark.Powell_at_eds.com says...
>
>

>> Then it is up to the developers who should know the data
>> better than the DBA to make the right choices.

>
>
> Are you suggesting that a DBA shouldn't have a notion about
> what data is in his databases, and not know why it is there and
> for what purpose it is being used?
>
> OK, maybe you didn't say "not have a notion" but you definitely
> said "should know the data better than the DBA".
>
>
> I disagree - a DBA should know more than application
> programmers about what's in the DB's he manages and why - how
> else can he make recommendations about how "things" could be
> done far better on a far far higher level than, perhaps, the
> programmers see the data? "Things" meaning "System
> organisation", "optimisation", &c. &c.

I'm not sure how a dba is supposed to do this. If the DBA is also the data architect, then, maybe. But then, the DBA will spend all his time exploring the business instead of architecting the database. I think it is a team effort. The DBA should be able to depend on a data architect to give him the needed information so he can make the decisions his application's datastore needs.

The developers are going to be the experts at what the sqlcode is doing and the DBA should be able to depend on them so he can help them right more performant code and also stay away from bad code.

I guess, nobody should think of themselves as the "sole" source of info.

-- 
Galen Boyer

Received on Tue Dec 16 2003 - 09:36:09 CST