Date: Mon, 15 Jul 1996 08:35:02 -0700
Reply-To: Karsten Self <kmself@IX.NETCOM.COM>
Sender: "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From: Karsten Self <kmself@IX.NETCOM.COM>
Subject: Re: Data Warehousing
For general warehousing concepts, I highly recommend Ralph Kimbal's "Data
Warehousing Toolkit". This discusses, in overview, most parts of the
warehousing issue. Kimbal discusses the "Star Schema" multidimensional data
model. SAS *deals* with lots of data, but does not offer the inherent
robustness of an RDBMS for creating and maintaining relational models.
I don't know your hardware, suggest you clarify the distinction between
optimizing for transactions processing (typical business system) vs.
analytic processing (typical warehouse) to your sysadmin or vendor.
At 09:16 PM 7/13/96 GMT, Bill Shannon wrote:
>I am biostatistician developing a data warehouse for the analysis of
>clinical data at a large midwestern academic hospital. I am going to use
>SAS for data management and statistical analysis and prefer a UNIX
>environment.
>
>I would appreciate any comments on peoples experiences, what pitfalls to
>avoid and what things I absolutely must do. I am especially interested
>in the following:
>
> Hardware options: I am considering a Sun Enterprise 3000
which I am
>told is optimized for I/O intensive operations.
>
> Database models: relational vs. multidimensional? How does
SAS do with
>multidimensional models?
>
>Please send anything evenly slightly relevant to me personally
>(shannon@osler.wustl.edu) or to the net. I will summarize what I
>receive. If you are in St. Louis and work on data warehouses please call
>me at 454-8356.
>
>
>Thanks
>Bill Shannon
>Washington University School of Medicine
>St. Louis
>shannon@osler.wustl.edu
>
>
----------------------------------------
Karsten Self / kmself@ix.netcom.com
...internet immortality achieved:
http://www.unitedmedia.com/comics/dilbert/puppets/HAMILTON.jpg
What part of gestalt don't you understand?
| 
