Re: How to pin a table in memory?

"Jonathan Lewis" <jonathan_at_jlcomp.demon.co.uk> wrote in message news:963445219.9584.0.nnrp-02.9e984b29_at_news.demon.co.uk...
>
> Not entirely true, especially since the CACHE
> option has only been around since about 7.1.
>

Yup. Ie, Year zero. (Poetic licence!)

> The CACHE option was to tell Oracle that
> a table that exceeded the small table threshold
> should nevertheless be treated as a small
> table for LRU/MRU purposes, provided it
> also fell below cache_size (default 10% of
> db_block_buffers).
>

Yup. And the enormous performance benefits from telling Oracle this are....?

>
> If the system is Oracle 8, then one option
> is to declare a KEEP buffer pool larger
> than the size of the table, and allocate
> the table to the KEEP pool. Once the table
> is pulled into the pool by non-scanning
> activity it will stay there.
>

And hence the CACHE clause is....

useless?

Regards
HJR
>
> --
>
> Jonathan Lewis
> Yet another Oracle-related web site: http://www.jlcomp.demon.co.uk
>
> Howard J. Rogers wrote in message <396c41bb_at_news.iprimus.com.au>...
> >The cache clause makes zero difference in the long run.
> >
> >Interesting that since year zero the Oracle training materials have
> >suggested you should 'use the CACHE clause for small, frequently used
> >tables'. To which the obvious rejoinder is.... if the table is
 frequently
> >used, where do you think the blocks from it are likely to be in any
 case??
> >
> >Short answer: it's a complete red herring, and a waste of time.
> >
> >Regards
> >HJR
> >
> ><Pebibit_at_pebibit.com> wrote in message
 news:396B214D.631785C4_at_pebibit.com...
> >> look at http://www.oracle.com/oramag/oracle/00-May/o30tun.html
> >> ALTER TABLE [TABLENAME] CACHE;
> >>
> >>
> >>
> >> CU
> >> pebibit_at_pebibit.com
> >>
> >> "Thomas Dreßler" wrote:
> >>
> >> > Hallo,
> >> > can anyone tell me, how to fix a table in memory?
> >> > DBMS_SHARED_POOL can only keep packages, triggers und procedures and
 the
> >> > CACHE parameter does'nt it.
> >> > Thomas
> >>
> >
> >
>
>
Received on Fri Jul 14 2000 - 00:00:00 CDT