Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Different results

Re: Different results

From: Frederic Mockel <frederic.mockel_at_gillardon.de>
Date: Wed, 21 Mar 2001 15:52:15 +0100
Message-ID: <3ab8bfd5$0$4943$4dbef881@businessnews.de.uu.net>

Ok,

we have written an application which execute delete-, insert- and update-statements against the db. The application use ESQL/C The commit comes all 1000 statements.

We run the application on the databaseserver in the night.The databaseserver is a HP-UX 10.20.

When the job is ready we take SQLPlus and run a query against the database to analyse the data.
(Normally we use a Win32-Application which execute that query as a dynamic-sql-statement).

The query is like that:

select t1.col1, t2.col2
from schem1.tab1 t1, schem1.tab2 t2
where t1.key = t2.key
union all
select t1.col1, t3.col1
from schem1.tab1 t1, schem1.tab3 t3
where t1.key = t3.key

We got a result for the first querie but we don't got a result for the second. Examining the data manually we should got a result for the second querie too.So the result is definitly wrong.

Then we use the export and the import-utility from Oracle. We make a schema-export, drop the schema an import the schema again.

When we now execute the statement we got a (right) result.

Why can this happen ?

Freddy

"Daniel A. Morgan" <dmorgan_at_exesolutions.com> schrieb im Newsbeitrag news:3AB84873.90148454_at_exesolutions.com...
> From your description I can not understand what you are doing. Can you
 clarify
> what you are doing and perhaps give some more details on the nature of the
 query
> and whether you are logging any errors?
>
> Daniel A. Morgan
>
>
>
> Frederic Mockel wrote:
>
> > Dear all,
> >
> > i have a problem.
> >
> > We work on Oracle 8.0.6.2 on HP-UX. We have very huge database ( 12 GB)
 with
> > few very big tables ( two tables have 43 million rows, the next on has
 10
> > millions rows).
> >
> > During the night we have a batch-job manipulating the data. We have
 updates,
> > deletes and massive inserts.
> >
> > To get the result we have a union-all-query.
> >
> > And now the effect:
> >
> > When we run the query against the database after manipulating the data
 we
> > got now result for the second statement of the query.
> >
> > When we do an complete reorg we got a (right) result for both statements
 in
> > the query.
> >
> > How can this happened ?
> >
> > Freddy
>
Received on Wed Mar 21 2001 - 08:52:15 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US