Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Usenet -> c.d.o.misc -> Re: Multiple datasets in one dataset
On Wed, 17 Oct 2007 23:52:40 -0700, roman.morokutti_at_googlemail.com
wrote:
>Hi Brian,
>
>if such sample data appear within a certain
>time span, we encounter a significant raise
>of time. In other words I quit the thread
>after about 4500 seconds.
What is the execution plan? In SQL*Plus use the command (i hope i get this right) "SET AUTOT TRACE EXP STAT", and run the query, then post the output here.
>After excessive testing it turned out, that
>those values were blamable where all values
>are 0.
Could those records bve ignored? If so, use a WHERE clause to exclude such a case, such as, AND A.Value1 > 0 AND A.Value2 > 0 ... AND A.Value10 > 0.
>We have also tested on Postgres ( before we
>turned it to an inner select - Postgres does
>not understand (+) ).
(+) is Oracle specific, just like *= is SQL Server specific. All support the ANSI syntax, however.
> Even the inner select did not return, either on Oracle nor on Post-
>gres.
An EXPLAIN PLAN would probably be helpful here as well.
B.
>Postgres moaned that memory exceeded. Oracle
>did not state anything. So this kind of data
>seems to be an impossible variation of data
>for this request.
>
>But as Postgres unveiled we may have to less
>dedicated memory for our Oracle database.
>
>Regards
>Roman
Received on Thu Oct 18 2007 - 08:54:03 CDT
![]() |
![]() |