Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Index compression vs. table compression

Re: Index compression vs. table compression

From: Richard Foote <richard.foote_at_bigpond.nospam.com>
Date: Tue, 04 Jan 2005 20:58:08 GMT
Message-ID: <ArDCd.103618$K7.3347@news-server.bigpond.net.au>


"Noons" <wizofoz2k_at_yahoo.com.au> wrote in message news:1104870566.505747.323500_at_f14g2000cwb.googlegroups.com...
> Richard Foote wrote:
>>
>> For the RECYCLE pool therefore, the best candidates are large,
> randomly
>> access tables where blocks are unlikely to be reused by other
> processes,
>> *not* simply FTS tables as is often stated by Howard and others.
>
> Another use is for application "temp" tables. By this I mean: in some
> instances (namely Peoplesoft apps), applications dump rows in
> intermediate tables with a userid on each row. These rows live
> for the duration of the transaction and are deleted at the end
> or by some background mechanism. Once used by a given transaction,
> the rows are essentialy useless for anyone else. It is therefore
> desirable that they use the least possible resources. RECYCLE is
> the correct cache for those tables.
> Yes we all know about temporary tables. But does Peoplesoft
> know about them?
>

Hi Nuno,

But if the *blocks* are quickly reused by new inserted rows from a subsequent transaction to replace the deleted rows because the pctused setting was set appropriately, then perhaps the RECYCLE is not the right pool.

It's the reuse of blocks not rows that's important.

Devil's advocate :)

Cheers

Richard Received on Tue Jan 04 2005 - 14:58:08 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US