Re: Wrong initial extent in XXX_SEGMENTS for LMT. Minor annoyance, will oracle ever fix this ?

On Thu, 8 Apr 2004 16:28:11 -0400, G Dahler <yellow-shark_at_spamex.com> wrote:

>
> "Howard J. Rogers" <hjr_at_dizwell.com> a écrit dans le message de
> news:opr55molwu3d8uqx_at_news.optusnet.com.au...
>> On Thu, 8 Apr 2004 15:45:36 -0400, G Dahler <yellow-shark_at_spamex.com>
>> wrote:
>>
>>
>> See. You knew it all along. Please don't use words like "ignore" when it
>> clearly, and you know it, doesn't "ignore" it.
>>
>
> Doh ! Sorry, sometimes words do not come right, I'm not english
> speaking, I
> meant that in the sense of the ACTUAL first extent that oracle allocates,
> that extent does not have the size you specified. By ignore, I meant,
> oracle does not respect what you tell it do do (It can't anyway, because
> of
> the LMT, but IMHO, it should/could put the correct value in the
> dictionary
> tables)

Well, I suppose I shouldn't assume that everyone speaks English (not even most English speakers do, actually), so my apologies.

But my point was that Oracle *does* respect your request for an initial request of X. It might not be able to give you precisely X, but it does its damndest to get there.

Sure, the extent sizes you get to make up near-X are not actually X. But that's not the same thing as saying it ignores your request altogether (as it does, utterly, for the next extent size, as you mentioned originally).

Subtle point, I guess. And maybe I just have too much time on my hands at 5.15am.

As to whether the data dictionary tables are incorrect, well... see that's where I disagree too. When you create a table, with a storage clause, DBA_TABLES (or DBA_SEGMENTS) correctly reports your create table statement specifications. The fact that what you see there is not what actually makes up the table is not DBA_TABLES/DBA_SEGMENTS fault. There's a difference between the logical definition of a table (which DBA_TABLES/DBA_SEGMENTS correctly reports) and the physical storage of that table (which DBA_EXTENTS correctly reports). The fact that users unfamiliar with LMTs (or forgetting what they're supposed to know about them) think the two are the same is unfortunate, but not Oracle's problem.

That's one way of looking at it, anyway. As I said, though, when DMTs are history, so will the confusion be, I expect. Regards
HJR

-- 
-------------------------------------------
Dizwell Informatics: http://www.dizwell.com
  -A mine of useful Oracle information-
          -Windows Laptop Rac-
    -Oracle Installations on Linux-
===========================================