Re: 30 Character Limit on Oracle

From: Richard Foote <richard.foote_at_bigpond.com>
Date: Mon, 17 Feb 2003 23:38:33 +1000
Message-ID: <8q44a.49783$jM5.124283@newsfeeds.bigpond.com>

"Harish" <harishparam_at_yahoo.com> wrote in message news:26f800fb.0302170401.23f180ef_at_posting.google.com...
> Hi,
>
>
> My Apologies if the question sounds a bit silly. I need to know the
> following
>
> 1. Oracle imposes 30 character limit on object names. Any specific
> reasons why this limit has not been relaxed even in 9i.

Hi Harish,

Can you image the following sql:

select * from
the_rise_and_fall_of_ziggy_stardust_and_the_spiders_from_mars_is_as_brillian t_an_album_now_as_it_was_when_it
was_first_released_over_30_years_ago_and_i_must_say_i_can_not_wait_until_the _ziggy_stardust_motion_picture_gets_released_on_dvd_later_next_month;

Can you imaging storing this in the DD or in the DD cache or in a procedure ...

30 chars is not such a bad limit.

>
> 2. SMON only coalesces contingous free extents when the pctincrease is
> set to non-zero value.
> What is the reason for this ?

Because if you have a pctincrease of 0, you *hopefully* might just have uniform sized extents making coalescing redundent. That said, with Dictionary Managed Tablespaces there's nothing to guarantee that. However, if you have a non 0 pctincrease, you're pretty well guaranteed fragmentation issues which coalescing *might* help to some degree.

My advice, give SMON a break and used Locally Managed Tablespaces where your question then becomes redundent.

Cheers

Richard Received on Mon Feb 17 2003 - 07:38:33 CST