Re: Locally Managed Tablespaces ... again!!!

"Jonathan Lewis" <jonathan_at_jlcomp.demon.co.uk> wrote in message news:b0ofm7$kgd$1$830fa17d_at_news.demon.co.uk...
>
> I think you and Howard (or Howard and you, there is no
> accusation intended in order of names, or apparent
> recipient of the mail) should just stop arguing.
>
> The demo statistics I posted, and your sample explain
> between them why the pair of you can so easily waste
> so much time over a trivial point
>
> >> I submit 960 samples to the DNA sequencing facility, so I supply
> the
> >> application with all the necessary data, and I hit the "submit"
> button.
> >> So 300 seqfac_todo blocks get dirtied, and 500 seq_fac_todo_idx_*
> blocks
> >> get dirtied, and these are next to each other in the cache list
> with maybe
> >> a few blocks from HR thrown in. No?
>
> You have a particular type of system - it is a little unusual to see
> something that looks like 960 rows changing that affect only
> 300 table blocks, but impact on 500 index blocks. The combination
> of row size, number of indexes, number of changes between null
> and not-null is obviously engineered for a specific, and non-generic,
> application.
>
> In one test case I showed that one table I/O could be matched
> by dozens of index I/Os - in which case you can forget about
> where the table goes and make sure you separate and spread
> the indexes as much as possible; in another case I showed one
> index I/O matching dozens of table I/Os - in which case you
> can forget about where the indexes go (even to the extent of
> putting them into the same tablespace as the table) so long
> as you smear the table across multiple devices.
>
>
> Bottom line:
> NEITHER of you is saying you shouldn't separate table
> from index.
> NEITHER of you is saying that if you've separated a table
> and it's indexes then the job is done.
> If you read each other's posts carefully you would notice
> that you are arguing at cross-purposes.
>

Not quite. He's arguing that there's an intrinsic propensity to contend on the writes. I'm saying it's not intrinsic. Sorry if we're boring you.

HJR
>
> Hmm. 99.73 OCR
>
> --
> Regards
>
> Jonathan Lewis
> http://www.jlcomp.demon.co.uk
>
> Coming soon a new one-day tutorial:
> Cost Based Optimisation
> (see http://www.jlcomp.demon.co.uk/tutorial.html )
>
> Next Seminar dates:
> (see http://www.jlcomp.demon.co.uk/seminar.html )
>
> ____England______January 21/23
> ____USA_(CA, TX)_August
>
>
> The Co-operative Oracle Users' FAQ
> http://www.jlcomp.demon.co.uk/faq/ind_faq.html
>
>
>
>
>
> ctcgag_at_hotmail.com wrote in message
> <20030121211936.349$h1_at_newsreader.com>...
> >"Howard J. Rogers" <howardjr2000_at_yahoo.com.au> wrote:
>
>
>
Received on Thu Jan 23 2003 - 05:15:43 CST