Re: Dictionary-managed tablespace

From: Paul Brewer <paul_at_paul.brewers.org.uk>
Date: Sat, 13 Jul 2002 20:24:40 +0100
Message-ID: <3d308905$1_2@mk-nntp-1.news.uk.worldonline.com>

"Richard Foote" <richard.foote_at_bigpond.com> wrote in message news:y3RX8.33999$Hj3.102708_at_newsfeeds.bigpond.com...
> Hi,
>
> Tag, my turn :)
>
> If all your extents are of the same size (because you set INITIAL = NEXT
and
> PCTINCREASE = 0) then you have *no* fragmentation (as all 'holes' left
> behind by dropped objects are totally reused). If you have no
fragmentation,
> you have no need for SMON to coalesce anything (basically SMON goes on a
> holiday to a nice Aussie beach and has a nice relaxing break :).
>
> PCTINCREASE = 0 is definitely the way to go and forget about SMON
> coalescing.
>
> Cheers
>
> Richard

Hi again, Richard.

Agreed. And (in the days before LMT), AFAIK, this was precisely *why* SMON didn't bother coalescing free space in tablespaces which had a default PCTINCREASE = 0.
So I've always wondered why there were some people recommending a tablespace default of 1 and tables created with same initial and next, and pcti 0, so as to 'fool' SMON and force it to try to coalesce periodically anyway. I have been assuming for years (without, I admit, having got around to testing) that this was yet another of those absurd myths.

Comments, anyone?

Thanks,
Paul Received on Sat Jul 13 2002 - 14:24:40 CDT