Re: BIG TABLES, small tables

Hi Chris,

    If I understand your question correctly I suggest that you do create the tables differently. I prefer to have three standard sizes; small, medium, and large. I make the medium tables ten times the size of the small and the large ten times the size of the large.

    When it comes to the number of extents I prefer to size them so that the initial extent is 25% of the anticipated maximum size. I don't think having several dozen -- or more -- extents for a table, index, or tablespace is a good idea.

Regards
Jerry Gitomer

Chris G. wrote in message ...
>How should these differ in terms of creation numbers? A table
with 200
>records that never grows much compared to tables that will grow
quickly to
>200,00 records and accessed a lot? Not concerned with indexes
or stats
>here. We have those.
>Rather, we now create all our tables with default numbers.
Should I create
>large tables differently than small tables? And if
>so, what numbers for a large table compared to a small table.
>
>Related issue: 150M datafile, with init 100k, next 100k, minext
1, maxext
>1400 pctinc 0.
> 1400 x 100 = 140000/1024 = 136.7M ??
>Do these numbers make sense for an idx tablespace? I gues what
I'm asking
>is the relationship with maxext to initial and calculating good
numbers?
>
>Thanks,
>
>Chris G
>
>
>
Received on Fri Sep 03 1999 - 10:59:39 CDT