Hans Forbrich wrote:
> Daniel Morgan wrote:
>
>
>>Hans Forbrich wrote:
>>
>>
>>>Caja wrote:
>>>
>>>
>>>
>>>>Hi there,
>>>>
>>>>I heard something about the automatic, configurable deletion of old
>>>>rows in tables in Ora 9i. This should take place in the CREATE TABLE
>>>>syntax.
>>>>
>>>>Does someone knows about that?
>>>>Thanks for any advice!
>>>>
>>>>Carsten
>>>
>>>
>>>I think you are looking for the Global Temporary Table definition.
>>>Either that or Advanced Queues (Oracle's name for message queues) which
>>>are implemented as using tables.
>>>
>>>Much more detail at htp://docs.oracle.com (and specifically for
>>>Oracle9iR2 at http://www.oracle.com/pls/db92/db92.homepage)
>>>
>>>If the former, look at the SQL Reference Manual for syntax and details.
>>>Also look at Concepts Manual and DB Administrator's Guide for necessary
>>>background and conditions.
>>>
>>>If Advanced Queueing, I suggest the following docco, in order:
>>>
>>> Concepts
>>> Application Developer's Guide - Fundamentals
>>> Application Developer's Guide - Advanced Queuing
>>>
>>>/Hans
>>
>>Sorry but no. I have checked my notes and I have an email from Mark
>>Townsend identifying that FIFO claim as relating to sorted clusters.
>
>
> I agree that my suggestions are not FIFO Tables which, as you say, are not
> implemented in 9i or before. Since OP didn't give details of the
> objective, I homed in on the 'automatic delete' that relates to GTTs,
> couched in a lot of 'I thinks, and ifs'. (AQ, on the other hand, provides
> FIFO and configurable autodelete, but not as traditional tables.)
>
> However, 10g docco has an interesting discussion and example around the
> Sorted Hash Cluster as related to a FIFO at
> http://download-west.oracle.com/docs/cd/B14117_01/server.101/b10739/hash.htm#ADMIN01901
> - nothing I've seen yet indicates it's got an auto-delete though.
>
> Then again, there is nothing [technically] stopping OP from developing a
> FIFO Table using Object Techniques, is there? <g>
>
> /Hans
You are correct. I have no idea where the autodelete idea originated.
First time I've seen it. My suggestion ... DBMS_SCHEDULER if someone
really wants it.
--
Daniel A. Morgan
University of Washington
damorgan_at_x.washington.edu
(replace 'x' with 'u' to respond)
Received on Tue Sep 21 2004 - 19:56:54 CDT
