Re: In an RDBMS, what does "Data" mean?

"Laconic2" <laconic2_at_comcast.net> wrote in message news:bKWdnZSDTJWsiVLdRVn-jw_at_comcast.com...
>
> "x" <x-false_at_yahoo.com> wrote in message
news:40cef875$1_at_post.usenet.com...
> > **** Post for FREE via your newsreader at post.usenet.com ****
> >
> >
> > "Laconic2" <laconic2_at_comcast.net> wrote in message
> > news:TPSdnXgqT7AJcFPdRVn-jA_at_comcast.com...
>
> > > When I built a "data mart" in Oracle as a star schema, I included all
> the
> > > primary and foreign key constraints, even though it slowed down
loading.
> > > The advantage came when I went to copy the star into Cognos (Impromptu
> or
> > > Power Play, I forget)
> > > Both Cognos and the Oracle optimizer recognized my star schema for
what
> it
> > > was, and made appropriate use of that fact.
> >
> > You were lucky.
>
>
> I don't think so. The engineers who built the CBO for Oracle were real
> smart. And they had the example of the DEC Rdb optimizer to guide them.
> And there was a note somewhere in the release notes saying they had
> implemented a thing they called a "star join". That was enough for me.

Is the star join a relational concept? I heard someone suggest that fact-dimension tables with star schema is bad design, but I forget the rationale for that and they seem to be very effective.

> And the Oracle DBA, who had plenty of experience with databases that ran
> like molasses, was amazed at the performance I got out of this beast.
> Especially when she looked at my code, and didn't find any "hints" and
> already knew that my tablespaces had nothing but default parameter
settings.
>
> It's amazing how many times you "get lucky" by just following simple,
sound
> design, and by keeping things "as simple as possible, but not simpler than
> that." In the few places where you end up with a performance problem, you
> can typically tune locally, without ripples spreading all over the system.
>
>
> The engineers who built the Cognos data extraction tool were real smart.
> And if they knew MDDB down cold (which they must have), then they almost
> certainly knew how to recognize a star schema when they saw one. They
way
> I knew that they knew was by looking at the SQL the Cognos tool used to
> extract the data from my star schema. Sure enough, they "got it".
>
> A lot of people in this business get a lot of bang for the buck by
assuming
> that "everybody but me is an idiot". I've gotten a lot of bang for the
buck
> by assuming just the opposite: "nobody in this business is an idiot. But
> everybody makes mistakes, and some of them are idiotic."

Good rule of thumb.

> (I occasionally call people "idiots". But that's just venting.)

and that's fine behind closed doors. I'm thankful that there is less of that in public on this list than there was when I started (I'm not sure now what to do with the balls I had to grow at that time, but pleased that I no longer need them to chat here ;-)

Cheers! --dawn Received on Tue Jun 15 2004 - 19:21:21 CEST