Re: Table fragmentation

Richard Foote wrote:

>Rare, generally a result of a poor table definition, but still
possible

Why do you say that? Isn't PCTUSED defined for the general OLTP case of rows coming and going, while the volume issue could be a normal result of a periodic or one-time purge? (Especially of things purged based on something other than they would be partitioned on).

Example: Inventory movement tracking table partitioned by warehouse, purged by date. Seems fairly vulgar to me.

jg

--
@home.com is bogus.
Science!
http://www.signonsandiego.com/uniontrib/20050112/news_1c12fertile.html