Re: Q:Good preformance

Another thing is that it sets how much percentage of a block will be used by data, and how much space will be left for updating that data.

If too much space is left for update (PCTFREE), more blocks will be used, thus making the table grow unnecessarily, take in account that tables are limited to 121 blocks (depending on the OS), if a table reaches it's limit you'll have to export and import it...

-- 


                                                \\\|///
Bruno Kamiche                                   ( o o )
+===========================================oooO==(_)==Oooo=======+


| bkamiche_at_sanmiguelind.com.pe                                    |

| pl9423292_at_ipae.edu.pe                                           |

| bkamiche_at_geocities.com                                          |

| Home Site : http://www.geocities.com/SiliconValley/Heights/9448 |

|                                           ooo0                  |

| From Lima - Peru                          (   )   Oooo.         |

+============================================\ (====(   )=========+
                                              \_)    ) /
                                                    (_/


JC <jcygan_at_netxpress.com> wrote in article
<3425d1ab.2907150_at_206.67.207.6>...
> On 19 Sep 1997 09:58:58 GMT, "Sanlam" <joe_at_soap.com> wrote:
> >you indicate how it was created and if there is contention. For inserts
 and
> >updates you have to look at the parameters on the tables PCTUSED,
 PCTFREE
> >as well as storage parameters for fragmentation.
> 
> I am still confused on PCTUSED, PCTFREE.  Could someone give me a good
> plain english explanation of how these influence performance.  I
> understand what row chaining is, and how bad it can potentially be.
> Is that the sole concern with these settings?
> 
> Joe
> 
>