Re: Convert SAP Oracle Database to IBM DB2 Database??

From: DA Morgan <damorgan_at_psoug.org>
Date: Sun, 10 Feb 2008 11:08:28 -0800
Message-ID: <1202670476.917879@bubbleator.drizzle.com>


Serge Rielau wrote:

> DA Morgan wrote:

>> Depends on the data. Have you looked at bank data? phone company data?
>> online reseller data? It isn't as random as you appear to think.
>> Consider, for example, the clustering factor on an index. That same
>> table that has address information to which you refer also has cities,
>> state/provinces, first names, and lots of other very predictably
>> repeating values. And, yes, at the block level.
> Are you saying that Firstnames strongly correlate with your clustering?
> Streetnames? What do you think is better:
> Store "Main Street" and "Smith" in a global dictionary or repeating it 
> in every block.

I am saying that while street names may be random, though often they are not. The use of North, South, East, West, Avenue, etc. is highly predictable. So are common first names and a lot of other information. One example ... in my city every phone number begins with 206-232 or 206-236.

>> What I am surprised you have not yet given voice to is the implication,
>> in Oracle, of using small blocks versus large blocks.
> You teach. I comment.

I know ... Mark can comment if he wishes. <g>

-- 
Daniel A. Morgan
Oracle Ace Director & Instructor
University of Washington
damorgan_at_x.washington.edu (replace x with u to respond)
Puget Sound Oracle Users Group
www.psoug.org
Received on Sun Feb 10 2008 - 13:08:28 CST

Original text of this message