Re: Best Practices
From: Michael Prendergast (mikepre_at_hotmail.com)
Date: 02/05/04
- Next message: Ching: "Connecting cubes in Analysis Server"
- Previous message: Michael Prendergast: "Aggregate Function in a stock measure"
- In reply to: Michael Morse: "Best Practices"
- Next in thread: Bluetooth: "Re: Best Practices"
- Reply: Bluetooth: "Re: Best Practices"
- Messages sorted by: [ date ] [ thread ]
Date: Thu, 5 Feb 2004 11:51:02 -0300
That will create an excesive work load over your transactional server.
Onother point, what will happen in one or two years when you cleanse your
tables looking for performance? Your views will also loose data.
The idea of a DWH is avoiding OLTP server overload with heavy analysis
queries and of course, maintaing your valuable data in the long term.
Hope it helps.
Michael
"Michael Morse" <mike.morse@micromo.com> escribió en el mensaje
news:2890C6D4-4CD1-439F-BA5D-E4B99B7D5508@microsoft.com...
> On our main SQL server I have setup a database called DataWarehouse which
is a location that through DTS all other databases dump relevant data for
analysis purposes. The DTS packages are doing ETL into, specifically,
tables in this DataWarehouse database.
>
> Would it be better if in this DataWarehouse database I used views instead
of tables to pull information from the tables I need? It seems like this is
easier than creating DTS packages to jump a new or changed data to the data
warehouse.
>
> Our main analysis is really on of one DataBase which is our ERP system. So
it's not like were dumping 5 different database into one location. It's
only 1.
>
> Thanks for the help
>
> -mike
- Next message: Ching: "Connecting cubes in Analysis Server"
- Previous message: Michael Prendergast: "Aggregate Function in a stock measure"
- In reply to: Michael Morse: "Best Practices"
- Next in thread: Bluetooth: "Re: Best Practices"
- Reply: Bluetooth: "Re: Best Practices"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
