Re: conventional path export with direct=y

Gerry Sinkiewicz wrote:
> "Frank van Bortel" <frank.van.bortel_at_gmail.com> wrote in message
> news:en8067$gho$1_at_news5.zwoll1.ov.home.nl...
> > joel garry schreef:
> > > This one is causing hair loss.
> > >
> > > hp-ux 11.11, oracle 9.2.0.6
> > >
> > > I have a script that has been running for a couple of years now with a
> > > series of direct path exports, no problems. I recently created a new
> > > schema that is a subset and ETL of an existing schema, through a series
> > > of imports and SQL modifications. The application that uses these
> > > schemata is a bit picky about its data, and doesn't seem to have a
> > > problem with any of the schemata.
> > >
> > > The exp commands in the script are of the form (each on one line):
> > >
> > > exp abcdef/xyz DIRECT=Y compress=N recordlength=65535
> > > file=/oradata/exports/abcdef.exp log=/oradata/exports/abcdef.log
> > >
> > > The result of this is an export with only about a quarter of the
> > > tables, and none of the other objects like views, stored procedures,
> > > triggers, etc. It just lists the tables that it exports, then says
> > > Export terminated successfully without warnings. It also says it is
> > > using conventional path!
> >
> > I know, not all tables can be exported with direct=y (LOBs mainly);
> > exp will automatically switch to conventional export (and note that
> > in the log). Also, NLS_LANG should be set
> >
> (http://asktom.oracle.com/pls/ask/f?p=4950:8:::::F4950_P8_DISPLAYID:46622033
> 5026)
> > but have never seen failure to do so export only a quarter of
> > the tables.
> > However, I never have used recordlength.
> > --
> > Regards,
> > Frank van Bortel
> >
> > Top-posting is one way to shut me up...

>

Well, since it worked for me when I cut and pasted to a regular old oracle unix login, I don't think it is a privs or LOB problem (I don't think there are LOB's, but it's been a while since I've checked). But I'll be checking to be sure it is running as oracle and poke around the data types anyways. I'm currently running the live load (which drops and recreates the schema), so hopefully that will fix the problem as mysteriously as it appeared.

I'm still looking for suggestions, anyways.

Thanks guys, and Happy New Year!

jg

--
@home.com is bogus.
"How about a staggeringly pompous, interminable, uninformed, dishonest
spasm of intellectual chicken-choking in a major magazine? How 'bout
that?"  - Sarah W. Gaymon