Re: Design question
From: Geoff Muldoon <geoff.muldoon_at_trap.gmail.com>
Date: Tue, 17 Feb 2009 09:17:50 +1100
Message-ID: <MPG.2404911bd2a6f2e8989716_at_news.x-privat.org>
Mladen Gogala says...
> On Mon, 16 Feb 2009 20:34:59 +0100, Matthias Hoys wrote:
>
> > Hello group,
> >
> > Oracle version is 10g. 11g might be possible too. I'm currently involved
> > with the pre-analysis of a new project I'll be working on.
> > We will get historical environmental data (temperature, rainfall, ...)
> > for about 70.000 different locations, and this for a period of 30 years,
> > with 1 measurement for each day of the year. So the total amount of
> > records will be around 750 million (70.000 x 365 x 30). This sounds
> > pretty much like a datawarehouse, no ? ;)
Date: Tue, 17 Feb 2009 09:17:50 +1100
Message-ID: <MPG.2404911bd2a6f2e8989716_at_news.x-privat.org>
Mladen Gogala says...
> On Mon, 16 Feb 2009 20:34:59 +0100, Matthias Hoys wrote:
>
> > Hello group,
> >
> > Oracle version is 10g. 11g might be possible too. I'm currently involved
> > with the pre-analysis of a new project I'll be working on.
> > We will get historical environmental data (temperature, rainfall, ...)
> > for about 70.000 different locations, and this for a period of 30 years,
> > with 1 measurement for each day of the year. So the total amount of
> > records will be around 750 million (70.000 x 365 x 30). This sounds
> > pretty much like a datawarehouse, no ? ;)
> Matthias, this looks pretty much like a classic star schema to me.
> Locations will, of course, be the fact table and measurements will be
> dimensions. You will need one table for each type of measure.
Shouldn't that be measurement as a fact and location as a dimension?
GM Received on Mon Feb 16 2009 - 16:17:50 CST