Yes, as it's a thesaurus, the import is done on existing tables... The error
is related to a constraint (CTXSYS.DR_UNIQ_THS_PHRASE) which exists both in
the Unix and the Windows schema. That's why I was assuming that ctxload
itself was taking care of the duplicated records, but I'm not sure about
that...
"mommy dba" <mommydba_at_netscape.com> wrote in message
news:vptk4f9hikhj83_at_corp.supernews.com...
> Are you importing into existing tables? If so, check to see if there is
> an additional constraint on the unix side. This error should exist on
> both systems. It is an index constraint. Look at DBA_INDEXES and see
> if there are additional indexes on the unix side. If not there, check
> DBA_CONSTRAINTS and find out the same thing. Don't look at the names -
> look at the count. If the names are system generated they will most
> likely be different.
>
> Teppicamon wrote:
>
> > Hi all
> >
> > We're experiencing a very weird problem importing a thesaurus through
> > ctxload. We have a thesaurus text file with some duplicated phrases on
it
> > (and yes, we know it, that shouldn't be like that but the file is really
> > huge and we'd rather avoid checking it again. Follow on, please...) and
we
> > make the import with Oracle 9.2 / Windows 2000. No problem, ctxload
seems to
> > know how to handle the duplicated records and the import ends correctly.
But
> > then we try to import on Oracle 9.2 / Unix. Then the file is parsed
> > correctly but when ctxload starts to insert we get this error:
> >
> > Beginning insert...DRG-50857: oracle error in drzmiip
> > ORA-00001: unique constraint (CTXSYS.DR_UNIQ_THS_PHRASE) violated
> >
> > Is that normal? Why this difference between Windows and Unix versions?
Is
> > there anything we can do to make the Unix ctxload behave like the
Windows
> > ctxload?
> >
> > Any idea will be very welcome...
> > ---
> > F.X. Gil
> >
> >
>
Received on Wed Oct 29 2003 - 02:49:11 CST
