Re: Novice question: Tablespaces and Users

From: Daniel Morgan <damorgan_at_x.washington.edu>
Date: Fri, 26 Dec 2003 12:26:21 -0800
Message-ID: <1072470288.370406@yasure>

Niall Litchfield wrote:

> "Richard Foote" <richard.foote_at_bigpond.com> wrote in message
> news:FpBFb.61770$aT.23137_at_news-server.bigpond.net.au...
> 

>>"Niall Litchfield" <niall.litchfield_at_dial.pipex.com> wrote in message
>>news:3fe6179d$0$28694$cc9e4d1f_at_news.dial.pipex.com...
>>
>>>Um
>>>
>>>See Daniel's answer. The idea behind his approach as I understand it is

> 
> to
> 

>>>partition tables on a client column and to store the partitions in
>>
>>separate
>>
>>>areas. This does do exactly as Daniel describes.
>>>
>>>It also proves my statement that " A user cannot own multiple objects of
>>
>>the
>>
>>>same type with the same name " to be woefully wrong.
>>>
>>>I ought to have known this
>>>
>>>
>>> 1 select object_type,count(*) from all_objects
>>> 2 where owner='NIALL' and object_name='SALES'
>>> 3* group by object_type
>>>SQL> /
>>>
>>>OBJECT_TYPE COUNT(*)
>>>------------------ ----------
>>>TABLE 1
>>>TABLE PARTITION 16
>>>
>>
>>Hi Niall
>>
>>I think you're being a bit hard on yourself. Your statement "A user cannot
>>own multiple objects of the
>>same type with the same name" is quite valid. In your example, there is
>>*only one* object that has a name called "SALES". However, SALES has 16
>>subobjects (table partitions) that belong to it, each identified by their
>>own subobject_name and object_id.
>>
>>With DBA_OBJECTS, Oracle tries to record the "parent" object that
>>corresponds to all partitioned objects within the same view. Therefore

> 
> *the
> 

>>same object_name* (parent) can be stored multiple times, but each

> 
> occurrence
> 

>>corresponds to a different subobject_name (child).
>>
>>Therefore the above example doesn't invalidate your correct comment. You
>>can't be blamed if DBA_OBJECTS decides to store the same object_name
>>multiple times.
>>
>>Cheers
>>
>>Richard

> 
> 
> 
> Thanks for the kind words. I have (to the extent that I have ever thought of
> partitioned objects i.e hardly ever) always thought of them as being 1
> logical object with several subdivisions (sub-objects) as you say. However I
> could also have done the equivalent as a select from obj$ with the same
> results. ISTM that if I have 16 objects in obj$ all of the same type, owned
> by the same user and all of the same name my original statement was just
> wrong. It doesn't mean that you can have 16 *tables* all of the same name,
> but Daniel's partitioning solution is still a perfectly feasible solution to
> the original problem. You can do mine on std edition for half to a third of
> the price though :(
> 

You are indeed being too hard on yourself.

Though, on the other hand, both packages and operators allow for overloading which are additional examples of multiple objects, in a sense, with the same name.

-- 
Daniel Morgan
http://www.outreach.washington.edu/ext/certificates/oad/oad_crs.asp
http://www.outreach.washington.edu/ext/certificates/aoa/aoa_crs.asp
damorgan_at_x.washington.edu
(replace 'x' with a 'u' to reply)

Received on Fri Dec 26 2003 - 14:26:21 CST