Having argued for conversion from SAS to Oracle in one context and against in
another, let me say a few words in the few minutes that I have left in the work
day. Oracle has two areas of unquestioned strength: - as an indexed server
supporting numbers of data entry and editing clerks who need to access small
sets of records; - as a server that maintains complex, indexed views of
relational databases. SAS has several areas of unquestioned strength: -
acquisition and editing of large files of data, particularly across platforms; -
descriptive and inferential statistics; - fuzzy searches through large volumes
of data; - ad hoc reporting from different sources of data. You will hear good
arguments from both sides on other capabilities and deficiencies.
If clerical or administrative personnel need access to an active database on a
Unix or NT server, I'd lean toward Oracle. If programmers or statisticians need
access to a database that is being updated in batch mode, SAS has the edge. Bad
database design can make either a pain to implement and maintain, but SAS and
good database design requires less overhead and works as well or better except
in Oracle's primary markets. Sig
From: joe17836@MY-DEJA.COM at Internet-E-Mail
Sent: Wednesday, July 12, 2000 7:59 PM
To: SAS-L@LISTSERV.UGA.EDU at Internet-E-Mail
Subject: RE: INSERT SASDATASETS into ORACLE Tables
In article <email@example.com>,
> Welcome to the wonderful world of the stellar Oracle performance! I
> is worth those M$$ that migrated to the Oracle coffers. On a
> note, please let me suggest that you should read the excellent SUGI
> by our fellow sasler Dianne Rhodes "Migrate to ORACLE? I need my SAS
> Software!"? Suffices it to say that its conclusion consists of but a
> sentence: "Stick with SAS".
> Kind regards,
> Paul M. Dorfman
> Jacksonville, Fl
Perhaps you know something about John's situation that I don't so my
comments should be read with that in mind. I will say I agree that
migration to Oracle is very expensive and, in many cases where the
business already uses SAS, unnecessary.
As far as proc dbload-ing data into Oracle goes, I have found that
dbload is rather inefficient as compared to, say, SQL*Loader which
should be able to handle loading 10,000,000 records in MUCH less time
than 24 hours. However, when pulling data out of Oracle, I have found
just the opposite -- SAS compares very well to and in some cases
performs better than other solutions such as "spooling", the PL/SQL
utl_file package, and Pro C.
Also, I am curious. Do you find that SAS can deal with very large
amounts of data as well as Oracle can? (I have no specific number in
mind when I say "large".) I have always been under the (admittedly
vague) impression that, for sheer speed in storing/processing very
large amounts of data, Oracle is superior. I realize that this all
depends on how the data is organized, indexed, etc.
Understand that I am asking these questions as a longtime user and
proponent of SAS. I am simply curious about your experience.
Sent via Deja.com http://www.deja.com/
Before you buy.